GOVERNMENT OF INDIA
DEPARTMENT OF ARCHAEOLOGY
CENTRAL ARCH/EOLOG I CAL
LIBRARY
CALLNe 05/
Acc. No. JLSJJSJL^
D.G.A. 79
GIP.\— S4— 2D. G. .■\rch. X. D./56.^25-9-58— l.OO.OiiO.
BULLETIN
OF
The School of Oriental Studies
BULLETIN
OF THE
-iCHOoL OF Oriental Studies
(UNIVERSITY OF LONDON)
Volume VIII ; 1935-37.
891-05,
•
-* 342 :
Published by
The School of Oriental Studies,
Agents: Messrs. Luzac & Co., 46 Great Russell Street, W.C. I .
CMiTR A L A " 5 ^OL.OGlGAi
LIBKAKV, *\l'w .
Aoe. No. /
Dftt*. J.9 '
Steihen Austin and Sons, Ltd.,
Pkinters, Hertford.
CONTENTS OF VOLUME VIII
PAPERS CONTRIBUTED
Dated Chinese Manuscripts in the Stein Collection. II. By
Lioxel Gile.s ........
The Japanese Particles 11 «. Ga. and Mo. Bv S. Yoshitake
S ome Corrections and Critical Remarks on Dr. Johan van Manen’s
Contribution to the Bibliography of Tibet. By Andrew
Vostrikov .........
A Fragment of the i'llaratantra in Sanskrit. By H. \V. Bailey
and E. H. Johnston .......
^ro Vedantic Hymns from the SuhlhdntamukltuxiU. By
Ananda K. Coomaraswamy
P6st(a). By fJARL CharpentIer .....
Bhagavata Purana and the Karikas of Gauclapada. By A.mar-
NATH Ray .........
Some Early Dramas in Bengali. By Jayanta Kumar Dasgi'pta
I ranian Studies. V. By H. W. Bailey .....
A History of the Caliphate in the Haijdt al-hcojau'dn of acl-DamIrl.
By Joseph de Somogyi .......
Nadir Shah's Campaigns in 'Oman, 1737-1744. By Laurence
Lockhart .........
The Arabic Theatre in Eg^'pt. I, II. By Nevill Barbour
G rammatical Categories in the New Hebrides. By A. Capell .
Bibliographif of the Published Writings of Sir George A. Grierson.
Compiled by Edith M. White ......
The Vasudevahindi, a Specimen of Archaic Jaina-Maharastri.
By L. Alsdorf ........
Yazdi. By H. W. Bailey .......
Does Khan Boll mean nothing more than Rustic Speech 1 By
T. Gr.ahame Bailey .......
Modern Literary Hindi. By A. Bar.\nnikov.
Remarks on the Pahlavi Ligatures and By K. Barr
S ur cpielf|ues dvandvas avesticjues. By E. Benveniste .
La Charrue vedicjue. Bv Jules Bloch .....
The Dialectical Position of the Niya Prakrit. By T. Burrow .
On the Alphabetic Notation of Certain Phonetic Features of
Malayalam. By Raymond T. Butlin ....
PAGE
1
27
51
77
91
101
107
113
117
143
437
VI
CONTENTS
PAGE
Sakadhuma. By Chaepentier 449
Purana Legends and the Prakrit Tradition in New Indo-Aryan.
By SuNiTi Kumar Chatterji . . . - • 457
Some New AwromanI Material prepared from the Collections
of Age Meyer Benedictsen. By Arthur Christensen . . 467
Les nasales en fin de mot en Sanskrit (et latin). By A. Cuny . 477
Der Typus tudd- im Altindischen. By Albert Debrunner . 487
The Prakrit underlying Buddhistic Hybrid Sanskrit. By
Franklin Edgerton ....... 501
Alphabets and Phonology in India and Burma. ByJ.R. Firth. 517
Mittelpersisch venok Erbse (Linse By Bernhard Geiger . 547
Singhalesische Etjuiologien. By IVilhelm Geiger . . . 555
The change of a to c in the Indo-Aryan Loan Words of Malayajam.
By K. Goda Yar.ma ....... 559
Observations on iliddle Indian Morphology. By Louis H. Gray 563
Sakische Etymologien. By Olaf Hansen .... 579
Soghdische Miszellen. By W. Henning .... 583
Der Tod des Kambyses : hvamnsyus amryata. By Ernst
Herzfeld 589
Bird-names in the Indian Dialects. By E. H. Johnston . . 599
Note on the Ancient North-Western Prakrit. By Sten Konow . 603
The Military Colonization of the Caucasus and Armenia under
the Sassanids. By J. H. Kramers ..... 613
Mala vihara. By fBYLVAiN Levi 619
Nochmals mleccha. By B. Liebich ..... 623
Nugae Burushaskicae. By D. L. R. Lorimer .... 627
Zur Schrift und Sprache der Kharosthi-Dokmnente. By
Heinrich Luders 637
Iranian Elements in Khowar. By G. Morgenstierne . . 657
The History of jp in Kanarese. By A. N. Narasimhia 673
Grammairiens tardifs et dialectes du prakrit. By Luigia Nitti . 681
The Expressions for “ The year consists of twelve months ’’ and
the like in Vedic Prose. By Hanns Oertel . . . 685
Syntax of the Past Tense in Old Rajasthani. By B. S. Pandit . 695
Yedico yiih “ se ipsum ". By Yittore Pisani . . . 699
Neue singhalesische Lautregel. By W. Printz . . . 701
Les noms de la moutarde et du sesame. By J. Przyluski and
C. Regamey 703
Sanskrit sd and sdh. By E. J. Rapson ..... 709
Pali hhunaha. By Baburam S.aksen.a ..... 713
Historia de Gramatica Concani. By JIari.ano Saldanha . . 715
CONTENTS
vii
Ein parthischer Titel im Sogdischen. By Hans Heinrich
SCHAEDER .........
On the “ Uralian ” Element in the D^a^^da and the MundS
Languages. By F. Otto Schrader ....
The Numerals in the Niya Inscriptions. By 0. Stein .
Tathagata and Tahagaya. By E. J. Thomas ....
/ Some Words found in Central Asian Documents. By F. W.
Thomas .........
Sanskrit a-hseti and Pali acchati in Modern Indo-Aryan. By
R. L. Turner ........
Some Dravidian Prefixes. By Edmtn H. Tuttle
Joan Josua Ketelaar of Elbing, author of the first Hindustani
Grammar. By J. Ph. Vogel ......
Altindische und mittelindische Miszellen. By J. Wackernagel
Wortkundliche Beitrage zur arischen Kulturgeschichte und Welt-
Anschauung. II. By Walther Wust ....
Two YazghulamI Texts. By I. I. Zarubin ....
^Ttaugara. By H. W. Bailey
. Hvatanica. By H. W. Bailey ......
/ Old-Iranian “ Peership ”. By Ernst Herzfeld
A New Contribution to the Materials concerning the Life of
Zoroaster. By S. H. Taqizadeh .....
The Copper Plate Grant of ^rivlfaraghava Cakravartin. By K.
Godavarma
A propos du genitif absolu en vied indien. By H. de Willman-
Grabowska ........
Dhola-Marura Duha : A Fifteenth-century Ballad from Rajputana.
Is the present conclusion original ? By T. Grah.ume Bailey
Phonetic Observations on the Brahul Language. By M. B.
Emeneau .........
PAGE
737
751
763
781
789
795
813
817
823
835
875
883
923
937
947
955
969
975
981
Ein tiirkisches Werk von Haydar-Mirza Dughlat. By Ahmet-
Zeki Validi ........ 985
The Arabic Theatre in Egj’pt. III. By Nevill Barbour . . 991
“ Ju-shih Lun ” — a logical treatise ascribed to Vasubandhu. By
Boris Vassiliev ........ 1013
A New Classification of the Constituents of Spoken Japanese.
By S. Yoshitake 1039
The Structure of the Chinese Monosyllable in a Hunanese
Dialect (Changsha). By J. R. Firth and B. B. Rogers . 1055
A Grammar of the Language of Florida, British Solomon Islands.
By W. G. IvENS 1075
vin
CONTENTS
PAGE
The Structure of a Bantu Language with special reference to
Swahili, or Form and Function through Bantu Eyes. By
E. 0. Ashton ........ HU
The e and o of LuGanda and the o of Swahili. By E. 0. Ashton . 1121
REVIEWS OF BOOKS
Liiujuistica. By R. L. Ttirner
1. Jules Bloch. L'Indo-Aryen du Veda aux Temps Moderiies 203
2. SuNiTi Kumar Chatterji. A Roman Alphabet for India . 212
3. M. Bloomfield. F. Edgerton. and M. B. Emeneau. Vedic
Variants. Vol. Ill ....... 213
4. Hjalmar Frisk. Zur indoiranischen und griechischen
Nominalbildung . . . . .214
5. Batakrishna Ghosh. Les Formations nominales et verbales
en j) du Sanskrit ........ 214
6. Rgveda-Samhita. Part I. Edited by Satischandra Seal and
Sitanath Pradhan ....... 215
7. Apastamblyadharma.sutram. Edited by G. Biihler. Third
edition .......... 216
8. The Unadisutras with the Vrtti of Svetavanavasin. Edited
by T. R. Chintamani ....... 217
9. Nandikesvara's Abhinaya-darpanam. Edited liy Manomohan
Ghosh ‘ , .217
10. The Padyiivall of Rupa Gosvamin. Edited by Sushil Kumar
De ■ ... 217
11. G. A. Grierson. On the Modern Indo-Aryan Vernaculars . 218
12. Pahudadoha of Ramasiriiha Muni. Edited by Hiralal Jain 218
13. Karakarhdacariu of Kanakainara Muni. Edited by Hiralal
Jain . . . . . . . , , .218
14. Jayadhavala Tika of Vlrasena. Edited by Hiralal Jain . 218
15. Krsnajanma of Maiibodh. Edited by Sri Umcsa Misra 219
16. Banarsi Das Jain. A Phonology of Panjabi as spoken about
Ludhiana and a LudhianI Phonetic Reader . 220
17. Srlmadbhagavadgita ka Pogrlbha.samivad : The Bhaga-
vadglta with the Text in Sanskrit and Translation in Dogrl.
By Gauri Shankar ....... 221
18. I. J. S. Taraporewala. Selections from Classical Gujarati
Literature. Vol. II. . . . . . 221
19. A. Darby. A Primer of the Marathi Language. Third
edition ......... 221
20. G. R. Hunter. The Script of Harappa and IMohenjodaro
and its connection with other Scripts. .... 222
CONTENTS
IX
PAGE
21. P. 0. Bodding. a Santa! Dictionary. Vols. II, III . . 222
22. Denys Bray. The Brahm Language. Parts II, III . 223
Bjorn Collinder. Indo-Uralisches Sprachgut. By Alan S. C.
Ross "... 227
K. Nielsen. Lappish Ordbok — Lapp Dictionary. Vols. I, II.
By Alan S. C. Ross ....... 234
Chandra-kanta Abhidhan. A comprehensive dictionary of the
Assamese Language. By P. R. Gurdon .... 235
Manichaische Handschriften der Staatlichen Museen Berlin.
Herausgegehen von Profes.sor Carl Schmidt. By H. W.
Bailey .......... 238
Lotus Delaporte. Syllabaire Hittite Cuneiforme. By H. W.
Bailey .......... 238
Caucasica. Fasc. 11. By H. W. Bailey ..... 239
Georges Dujiezil. Etudes comparatives sur les langues
caucasiennes du nord-ouest (Morphologic). By H. W.
Bailey 239
Textes populaires Ingus. Recueillis par M. Jabagi. Traduits
par G. Dumezil. By H. \V. Bailey ..... 239
Arthur Christensen. Die Iranier. By H. W. Bailey . 240
H. A. Jaschke. a Tibetan-English Dictionary. By H. W. Bailey 240
Reyue des fitudes Islamiques. Tome VIII. By H. A. R. Gibb . 241
The Holy Quran. English Text ami Commentary (with Arabic
Text) by A. Yusuf All. Part 1. By H. A. R. Gibb . . 242
La Sagesse Coranique eclairee par des Versets Choisis. Par
Mahmoud Mohtar-Katirjoglou. By H. A. R. Gibb . . 242
Rene Grousset. Histoire des Croisades et du Royaume Franc de
Jerusalem. II. By H. A. R. Gibb . . ' . .243
Book of Treasures. By Job of Edessa. Edited and translated by
A. Mingana. By A. S. Tritton ...... 250
Das Buch der Alaune und Salze. Herausgegehen, iibersetzt und
erlautert von Julius Ruska. Bv Paul Kraus . . 252
Ibrahlm Madkour. La Place d'al Farfibi dans I't^Je philo-
sophique musulmane. By Alfred Guillaume . . . 253
M. IsHAQi^E. Sukhanvariin-i-Iran dar ‘asr-i hazir. Poets and
Poetry of Modern Persia. Vol. I. By V. Minorsky . 254
C. A. Storey. Persian Literature. Section II. Fasc. I. By V.
IMinorsky ......... 255
Dialogues in' the Ea.stern Turki Dialect on Subjects of Interest
to Travellers. Collected and edited by .Sir E. D. Ross and
Rachel 0. Wingate. By V. IMinorsky . .... 257
X
CONTENTS
PAGE
Prosateurs Turcs contemporains. Extraits par E. Saussey. By
V. IVIinorsky ........ 259
Seyhi divanini tetkik. Part I. By Dr. Ali Nihat. By V. llinorsky 259
Ella K. Maillart. Turkestan Solo. By V. Minorsky . 260
Georges le Fevre. An Eastern Odyssey. By V. Minorsky . 261
Sir Percy Sykes. A History of Exploration from the Earliest
Times to the Present Day. By V. Minorsky . 262
Mary Countess of Minto. India, hlinto and Morley. By H.
Dodwell ......... 263
Edward Thompson and G. T. Garratt. Rise and Fulfilment
of British Rule in India. By H. Dodwell .... 264
L. S. S. O’Malley. India’s Social Heritage. By H. Dodwell . 265
Raymond Schwab. Anquetil-Dnperron ; SaVie. By H. Dodwell 265
L. W. Ly'de. The Continent of Asia. By H. Dodwell. . 266
Sir Charles Fawcett. The First Century of British Justice in
India. By S. V. FitzGerald ...... 267
The Vyavaharamayukha of Nilakantha. Translated into
Enghsh by P. V. Kane and S. 6. Patwardhan. By S. V.
FitzGerald ......... 269
S. Yoshitake. The Phonetic System of Ancient Japanese. By
Arisaka Hideyo ........ 271
Les Notes de Chevet de Sei Shonagon’, Dame d’Honneur au
Palais de Kyoto. Traduction par Andre Beaujard. By S.
Yoshitake ...... ... 275
Andre Beaujard. Sei Shonagon’, son Temps et son QEuvre.
By S. Yoshitake ..... ... 275
Sinica. By E. Edwards
1. Lin Yutang. My Country and my People . . . 1125
2. C. P. Fitzgerald. China. A Short Cultural History . . 1126
3. Lancelot Forster. The New Culture in China . . 1126
4. Kenneth Saunders. A Pageant of Asia .... 1128
5. Kenneth Saunders. The Ideals of East and West . .1128
6. Henning Haslund. Men and Gods in Mongolia (Zayagan) 1129
7. Reginald F. Johnston. Confucianism and Modern China 1129
8. Arthur Waley. The Way and its Power . . . . 1130
9. Luther Carrington Goodrich. The Literary Inquisition
of Ch’ien-Lung 1131
10. The Romance of the Western Chamber. Translated by S. I.
Hsiung 1132
CONTENTS
xi
PAGB
11. Tun Li-Ch'en. Annual Customs and Festivals in Peking.
Translated by Derk Bodde ...... 1132
12. CoRiNNE Lamb. Chinese Festive Board .... 1133
13. Chinese Art. Edited by Leigh Ashton .... 1134
14. Chinese Art. An Introductory Handbook . 1134
15. SoAME Jenyns. a Background to Chinese Painting . . 1134
16. Chiang Yee. The Chinese Eye ..... 1134
17. Arnold Silcock. Introduction to Chinese Art . . 1134
18. The Chinese on the Art of Painting. Translations and Com-
ments by Osvald Siren . . . . . . .1137
19. Benjamin March. Some Technical Terms of Chinese Painting 1138
20. Frank Davis. Chinese Jade ...... 1139
21. J. J. Brandt. Modern Newspaper Chinese . . 1139
22. Walter Brooks Broitner and Fung Yuet Mow. Chinese
Made Easy ......... 1140
23. J. B. Davidson-Houston and R. V. Dewar-Durie. Chinese
and Enghsh Modern Military Dictionary . . .1141
24. Lucy Driscoll and Kenji Toda. Chinese Calligraphy . 1142
H. W. Ponder. Cambodian Glory. By C. 0. Blagden . 1143
R. S. Wauchope. Buddhist Cave Temples of India. By F. J.
Richards ......... 1145
Ramaprasad Chandra. Medieval Indian Sculpture in the British
Museum. By F. J. Richards ...... 1146
L. S. S. O’Malley'. Popular Hinduism. By W. Sutton Page . 1147
Dhola-Marura Dtiha. Edited by Ram Sih, Suraj Karan, and
Narottam Das. By T. Grahame Bailey .... 1147
Tulsi Das’s Ramcaritmanas (Ramayan). Edited with Com-
mentary by Ram Nares Tripathl. By T. Grahame Bailey . 1148
Ziya Muhammad. Yadgar i Varis. By T. Grahame Bailey . 1149
Kastur Mal Bathiy’.y. Nama-lekha aur Munibl. By T. Grahame
Bailey .......... 1150
Shyam Sundar Das and Padm N.yray'an Ac.yry'a. Bhasa
Rahasya. Part I. By T. Grahame Bailey. . . 1151
The Jasmine Garland (Kundamala). Translated by A. C.
Woolner. By C. A. Rylands ...... 1152
The Meghaduta of Kalidasa. Translated by G. H. Rooke. By
C. A. Rylands ....... . 1152
E. D. Perry. A Sanskrit Primer. New and revised edition. By
C. A. Rylands ........ 1153
Codices Avestici et Pahlavici Bibliothecae Universitatis Hafnien-
sis. Vol. V. By H. AV. Bailey . . . . .1153
XU
CONTENTS
PAGE
Ernst E. Herzfeld. Archaeological History of Iran. ByH. \V.
Bailey .......... 1154
George C. Cameron. History of Early Iran. By H. W. Bailey . 1155
E. Benv’ekiste. Les Infinitifs Avesticjues. Bv H. ^Y. Bailey . 1155
Hannes Skold. Materialien zu clen Iranischen Pamirsprachen.
By. H. W. Bailey ........ 1156
Sarndliinirmocana Sutra. Texte Tibetain edite et traduit par
Etienne Lamotte. By H. \V. Bailey . .... 1157
Akhbar ar-Radi wal-Muttakl from the Kitab al-AyTak. Edited
by J. Heyworth Dunne. By Rhu von Guest . . . 1157
R. Blachere. Abou T-Tayyib Al-Motannabi. BvH. A. R. Gibb 1160
Francesco Gabrieli. II Califfato di Hi.sham. ByH. A. R. Gibb 1161
G. Bergstrasser's Grundziige des Islamischen Rechts. Bear-
beitet von Joseph Schacht. By H. A. R. Gibb . . 1163
Revue des Etudes Islamiques. Tome IX. By H. A. R. Gibb . 1164
Salomon Pines. Beitrage zur Islami.schen Atomcnlehre. By
A. S. Tritton ........ 1165
A. Heschel. Die Prophetic : 0 Istocie Proroctwa. By A. S.
Tritton . 1166
Traites Religieux. Philosophiques et Moraux, Extraits des
CEuvres dTsaac de Ninive. Edited by R. P. Paul Sbath.
By A. S. Tritton ........ 1167
Margaret Smith. An Early Mystic of Baghdad. By A. J. Arberry 1168
The Doctrine of the Sufis. Translated by A. J. Arberry. By
Margaret Smith ........ 1169
ShihabuddIn SuHRAWERDi M.iQTUL. Three Treatises on
Mysticism. Edited and translated by Otto Spies and S. K.
Khatak. By JIargaret Smitli ...... 1170
Sayyid Ahmad KasravI. Tarlkh-i jiansad sala-yi Khuzistan.
By V. Minorsky ....... 1172
A. Cremona. 11-Fidwa tal-bdiewa. By C. L. Des.soulavy . . 1175
DunKarm. UgoEo.scolo. L-0(jbra. By C. L. Dessoulavv . 1176
P. P. Saydon and Jos. Aqcilina. Ward ta' Qari Malti. By
C. L. Dessoulavy ........ 1176
A. Y. Vassaixo. Dell ta' 1-inighoddi. By C. L. Dessoulavy. 1177
A. il. Galea. S. Guann Bosco. By C. L. Des.soulavy. .1177
Dun Karm Farrugia. Taghlim Nisrani fuq il-Katekismu tal-
Kardinal P. Gasparri. By C. L. De-ssoulavy . . 1178
Gabra ta’ Ward. By C. L. Dessoulavy 1180
Edmund Sutcliffe. A Grammar of the Maltese Language. By
C. L. Dessoulavy ....... 1180
rOXTENTS XU
PAGE
Ida C. Ward. An Introduction to the Ibo Language. By
Margaret M. Green ....... 1188
Clemext M. Doke. Bantu Linguistic Terniinology. By E. 0.
Ashton ......... 1190
F. JoHxsox. Kamusi ya Kiswahili yaani Kitabu cha Maneno ya
Ki.s\vahili kinietungwa. Bv E. 0. Ashton . . . 1192
Ten African.?. Edited by Margery Perhain. By I. C. Ward. . 1193
Notes axd Queries ........ 1195
Obituary —
Profe.ssor Alice M’erner ....... 281
Professor Jarl Charpentier ... . . 282
Professor Basil Hall Chamberlain ..... 284
Professor Zoltan Gombocz ...... 286
Summaries of These.?. ....... 287
Books Eeceived for Review . . _ . . 294, 1199
BULLETIN
OF THE
SCHOOL OF ORIENTAL STUDIES
PAPERS CONTRIBUTED
Dated Chinese Manuscripts in the Stein Collection
By Lioxel Giles
(PLATE I)
II. Seventh Century a.d.
S OOX after the rise of the T'ang dynasty, the Parinirvana Sutra
seems to have lost its popularity, and in the second half of the
seventh century its place is definitely taken by the Lotus Sutra. Out
of a total of fifty-six dated texts, no fewer than twenty-nine are
sections of this work. The appearance of Taoism in the shape of three
hitherto unknown “ sutras ’’ (fo borrow a convenient Buddhist term)
is interesting but not surprising under a dcTiasty which extended
special favour to that ancient and much-corrupted cidt. It is hardly
possible to dignify with the name of religion such a strange medley
of magic, legend, and gross superstition ; and one cannot believe that
its scriptures were regarded very seriou.slv by any large section of
the community.
Altogether, the old atmosphere of simple faith and piety which
was so noticeable in the earlier centuries is now becoming more
attenuated. The introduction of the tabulated colophon (see year
fill and following) is symptomatic of this decline in true religious
spirit. Here, in place of the prayer on behalf of a deceased relative
or a living sufferer, or an unselfish offering made in the interests of
sentient existence as a whole .v^find ^nljr a 4i£^l enumeration of the
persons, lay or clerical.,,T(?s^mible fift manu-
script roll. The religioss element seems to Iiav^^^^^^Nimeezed out.
VOL. VIII p.cRT 1. ' AaN 1
No
o
LIONEL GILES —
This process continues in the ensuing centiuies ; for. though the
number of dated documents shows a large increase, the old type of
colophon becomes steadily rarer.
A.D. 601 (Sui).
S. 2048. ^ pk ^ ^ ^ — *She lun chang, ch. 1. The title
appears in the middle of the colophon, and is written in the same
hand. There are three different translations of the Mahayana-
sanigraha-sastra (X. 1183. 1184, 1247) and two commentaries on them
(X. 1171). Though the present work is styled a commentary (gft)
in the colophon, it appears to be rather an outline or resume of the
sastra.
Colophon: {n # ^ ^ A -tk A B JS. M ^ M ^
^ H ^ M ^ ^ ^ ^ Bit gfe M ^ iic [fo”
ftl Jt ^ -Sth day of tbe 8th moon of the 1st
year of Jen-shou [30th September, 601] the novice Shan-tsang of
the Ch'ung-chiao ilonastcry at Kua-chou copied the She hi)! .sa
in the Pien-ts'ai Monastery at tlie capital [Cli'ang-an] in order
that it might be circulated to the end of time. AVord-for-word
revision completed.”
Kua-chou is the modern An-hsi, east of Tunhuang. X'either of
these monasteries is mentioned again in the Htein ALSS. 5^ ” the
end of the dynasty ” is understood to be synonymous with the end
of time. Alas for human anticipations ! The 8ui dvnastv was
destined to last onlv 17 rears lonuer. The roll, made of thin golden-
yellow paper, is about 34 ft. long. 28 cm. wide. The handwriting is a
tine specimen of cursive or .semi-cursive.
602 (Sui).
S. 354S. rf* I'iiJ ^ *Chiu!<i a hit)) chiag (Aladhyamagama-
siitra,), ch. S. pin 4, nos. 2 (end only) -4. This corresponds to X. 542
(3.3- 35). K. .\ii. 9. The number of characters in the sections diff-u^
slightly from the figures given in tlie inoderu recension.
Colophon: t # r ^ -h - /] di- FI u ^ :r 'ib
m m it li. m m n p‘i ft m ^
P'f tk i) ” Copied by the .scribe Chang Ts'ai on the 20th day
of the 12th moon of the 2nd year of Jen-shou [6th Februarv, 602 1.
25 sheets of paper used. [Only 10 complete sheets remain.]
Revised by Seng-kai, priest (.sranuma) of the Ta-hsing-shan
-Monastery. Again revised by Fa-kang, priest of the Ta-chi
AI'jna.sterv.‘''
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION
3
The life of Seng-kai 'will be found in Hsil hto seng chuan, ch. 26 ;
K. XXX. 2. f. 177 r°. The Ta-hsing-shan Monastery is mentioned in
the colophons of several other Stein MSS. is used here for ^ or
m m)-
This is one of the finest MSS. in the Collection, the handwriting
lieing as nearly perfect as one can imagine. The roll is composed of
thin, crisp, golden-yellow paper of excellent quality, and measures
about 16 ft. by 26 cm.
603 (SxH).
S. 4553. M ^ J: J'O'fl Ruling ching. ch. 1
(out of 2 or 3). This is reallv a Buddhanama sutra. though the title
does not seem to suggest it. and probably apociA'phal. Contrary to
the usual practice, the invocations are not numbered.
Colophon : v’c [iff t ^ H T1 “h IS R fn ft ^
^ m t ^ M f'l ic ^ ^ m 'U -1] m.
IM — M ^ tli: # a Tt ^ #
hTr ^ ± ® ft fjfl |ft # IS +0 ft iS ^ B
ft .IE 4: M 5c mV % m R ^ ^ i'i iSf m <»
m m -k \i ^ ^ m it W} % w <* it ti .6, g
^ jg # -6, m H . 1 - k M m k -6 a >lf- I 5 H Ifi] _h il
” On the 14th day of the 2nd moon of the 3rd year of Jen-shou
in the Great Sui dynasty [31st March. 603] the female devotee Ling-hu
Fei-jen. having with pious intent cut off a jiortion of her expenditure
on clothes and other goods, has reverently caused a section of the
Ta shhuj fang hnang ching to be copied, praying that as a consequence
her parents of seven previous incarnations, the parents who begot
her in this life, and her faiuilv now living, wheresoever they are reborn,
may meet Buddha and hear his Law. falling in with good friends and
associates : and that after tlieir span of earthly re-incarnation is
completed they may be reborn in th.e Kingdom of limitless longevity
in the IVestern llegions : and that all living beings of the universe
may be included in tlii^ prayer. Also otfered by the female devotee
.len-sliih-shih. who gave help towards the copying as a means of
conversion to the faith. Offered with steadfast heart by Fei-jen’s
son Lii Sheng-tsun. Offered tvith constant heart by her daughter
A-man. who also pravs in the same sense as the above prayer.
It appears that Ling-hu tvas the donor's maiden name, while that
of her husliand was Lii.
This is a verv good 318. on thin yellow paper, rather discoloured.
The roll is over 18k feet long. 25 cm. wide.
4
LIONEL GILES —
604 (Sui).
S. 4162. ® ^ ^ M Yu p'o sai chieh cMng (N. 1088), ch. 2.
p in 10-12. There is no penera! title at the beginning.
Colophon: iz w US ^ m B A B W. Mi B t: X
i& it — s s If PpI iis m It # m “ li
it # ;1L Pi [for fi # ^ ^{r _ ^
■■ On the 8th day of the 4th moon of the 4th rear of Jen-shou
[llth May. 604] Ting (0 \\ ei-chen. on behalf of hi.s deceased father,
caused sections to be copied of the Yii p'o sai ching. the Kuan ting.
the Shan o gin kuo. the T'ai tzu ch'eng tao. and the TTw po wen shih :
he has also had made an image of Avalokitesvara. and a 49 ft. banner,
to the end that all living beings of the universe mav eventuallv achieve
Buddhahood."
^ is not a recognized surname, and it is almost certainly a mistake
for ^ : see the next entry. A'urta tmfjr is doubtless X. 167. Shan o yin
kuo ching is an apocrt'phal sutra of which a dozen copies are preserved
in the Stein Collection : see Ky5to Supplement. A. i. 4 (e). T‘ai tzu
ch‘eng tao {" The Crown Prince attaining enlightenment ") is a life
of Sakyamuni Buddha. TTm po wen shih means “ Five hundred
subjects of inquiry ” and is e\’idently a doctrinal work.
Another very good MS. on golden-yellow paper. The roll is
19 feet long, 26 cm. wide.
604 (Sui).
S. 4570. @ ^ ^ *Ya p o sai chieh, ch. 6. This isp'in 24 (1)
of N. 1088. It has a colophon very similar to the preceding, and of
exactly the same date : t S’ W PR TJ A H ^ t
X 'U m M ~ UK.fi PpI m ~ (jb t m ^
7t m ig ittf A f'i i)S — □
On the 8th day of the 4th moon of the 4th year of Jen-shou
Tang Mei-chen. on behalf of his deceased father, caused a section of
the Kuan tiiuj to be copied, a section of the TI a jm wen shih, and the
Fifteen hundred names of Buddha ; he has also had made an image
of Avalokitesvara, and a 49 ft. banner, to the end that all living
beings of the universe may eventually achieve Buddhahood.”
Comparing this with S. 4162. we ob.servc that the list of works
copied is shorter, but includes one new text ; and that the surname
of the donor is given as ^5 Yang. It is difficult to account for these
discrepancies, seeing that the two rolls were actually copied on the
same day. The handwriting of the colophons is the same, and probably
DATED CHINESE IIANUSCRIPTS IN THE STEIN COLLECTION
5
that of the texts as well, though it is somewhat smaller in the present
roll, which is only 4| feet long and 24-5 cm. wide.
606 (Sui).
S. 2598. ^ US M *Tii pan nieh p'an ching, ch. 16. This
corresponds to N. 113, K. viii. 5, from about the middle of ch. 16
to a point near the beginning of ch. 17. There is a touching colophon,
which is not very easy to translate : # -A~ ^ ^ 'Mg,
mm. m
“ In the year, the 2nd of Ta-yeh
in the Great Sui dynasty [a.d. 606], the bhikfshu Shih Shan-tsang. on
behalf of his deceased mother the Lady Chang, has reverently caused
a copy of this sutra to be made for circulation, as an act of worship.
Overcome with feelings of desolation and bitter grief that cannot
be assuaged, full of painful yearnings that can never cease, he ventures
to seek help from the great compassion of Buddha, that the latter may
put forth his limitless power. He prays that the virtue of these few
written characters may bring expectation of happiness to his mother
among the shades, and that the merit acquired by half a gatha may
help her along the road of spirituality. May the sound of the Law
never fade away, may the Yii-ling(i) be constantly handed down.
Though the fire in the kalpa [of destruction] be fierce, may the Dragon
Palace still abide. May the beings produced through the four kinds
of birth in the six paths of existence all alike exhibit surpassing
karma."
IVe have already met with Shan-tsang in the colophon to S. 2048
(a.d. 601) ; he was then a novice, now he is a full-fledged bhikshu.
Z1 ^ and ^ are instances of meiosis ; ho is evidently referring
to the section of the sutra that has just been copied, though it happens,
curiously enough, that there is no gatha at all in ch. 16. ^ " Fish
Ridge ’’ contains some allusion that I have not succeeded in tracking
down. the kalpa of destruction " is the third of the four
periods of universal evolution and decay. The three major calamities
which then appear are fire, flood, and wind. The Dragon Palace is
the abode of the Dragon King at the bottom of the great ocean ; it
was once visited by Buddha, who preached the Law there. P9 ^ is
a pleasing variation of the usual ^ ^ ; the four kinds of birth are
from womb, egg. damp, and metamorphosis.
LIONEL GILES —
&
This is a fine regular MS. on golden-yellow paper, about 19 feet
long.
608 (Sui).
S. 2419. ^ ® *Miaof<i lien hua ching (Saddharma-
pundarlka-sutra). eh. 3, p‘in 5. 6.
Colophon: H PS + S 0 SP Jg BiU
^ til M iS " Oil 15th of the 4th moon of the 4th
year of Ta-yeh [3rd June. 608] AVang Shih. Lu-shuai of Tun-huang
Chiin, reverently had copies made of single sections from the Nieh
p’an, Fa hua. and Fang kuang on behalf of his deceased mother, in
order to confer happiness on her honoured soul through this act ot
surpassing merit, and praying that she may avoid the three lower
paths of sentient e.vistence and rise to the seven states of purity :
and that all being.s endowed with perception in the si.x paths of
transmigration may be steeped in the ocean of pious vows
(pranidhana)."
-\t first sight. gjji J gjp would .seem to be two names (Master
Lu and blaster Wang) ; but I think it is more likely that the first
m ha s been mistakenly written for 6i|J- Ml being the title, created
under the Sui dynasty, of a high military officer known until then a>
HP ^ Tz'ii jiiidn. 193c. The (or 0) are the
seven characteristics of bodhi : di.scriniination of true and false,
zeal. jo^Tulness. discarding of all gro.s.sne.ss of body or mind, power
of remembrance, power to keep the mind in a given realm undiverted,
indifi'erence to all disturbances of the sub-conscious mind.
This is a good MS. on light yellow paper. ov('r 13.1 fi'et long.
612 (Sui).
S. 2295. /fcc/o : ^ ^ it *Coo f:/t pien Inm ching.
This ■■ sutra on the Incarnations of Lao Tzu ' is our first dated Taoist
te.xt. It is not included in the existing Taoist Canon.
Colophon: M ^ fi +119 0 I ii ^ "
f]] ra ^ ^ HIS iSi iii ± ® O ^ A <>
Copied by the .scribe AVang Ch ou on the 14th day of the 8th
moon of the 8th year of Ta-yeh [ 1 4th September. 612]. Four sheets of
paper u.sed [showing that only a few columns are wanting to our
roll]. Again revised by a Taoist priest of the Mystic Temple in the
Mystic Capital. Paper-dyer. [ ]. Copied in the Secret Books
Department."
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION
7
For cf. S. 3548 (a.d. 602). The use of ^ (dark, mysterious,
profound) is affected by Taoists for anything appertaining to their
religion. Here no more seems to be meant than a Taoist temple at
the capital. Ch‘ang-an. For ^ cf. T‘ang shu, ch. 47, f. 3 v° ;
^ ^ ^ E A A “ Pight makers and dyers of paper."
According to the Tz'u yiian. the term is now applied to the mounting
of scrolls and pictures. The name of the dy'er has been omitted here.
Verso (in an inferior hand) : Extracts from other Taoist, or semi-
Buddhist, texts.
This i.s a fine MS. on yellow paper, 6f feet long.
616 (Sui).
S. 2605. ^ ^ H'! llx ^ II *Fo shno chin hnig pan jo
ching. This is the end jiortion only of Kumarajlva's translation of
the Diamond Sutra (N. 10). The usu.il form of the title is ^ ^i]
^ -iUi n m-
Colophon: A 1^ A H + “ A /I il' - H ft ^
m ^ M M Wc M Wc ^ m n ~
[? for #] S ^ X M ill # ^ SI ^ m it M ^ p
#15 Is -'5rd day of the 7th moon of the
12th year of Ta-yeh in the Great Sui dynasty [9th September. 616]
the female lay devotee [upusika] Liu Yiian-ching reverently caused
a copy to be made of this sfitra. praying that by this slight act of
merit, M'hen it i.s read aloud on behalf of all living beings, all those
who hear it may reverently believe, and awaken to the unreality of
suffering, and that those who read it may hold fast to its precepts
and all ri.se to enduring felicity. She also prays that she herself
may soon quit this desert frontier region and speedily return to tin*
Imperial capital ; that the barriers of her sin may be broken down,
and that manifold blessings may descend upon her."
This fragment, of thin light yellow paper, is not more than 2J feet
long. The text and colophon are in the same hand. It i.s our first
dated example of the Diamond Sutra, which in Kumarajlva’s trans-
lation was to be by far the most popular of all the short sutras pre-
served at Tunhuang. Yet out of several hundred manuscripts only 21
are complete.
626 (T'.wo).
S. 4635.
/c/i la shan fan pa ch'iieh hsing shift ch'ao. ch. 1 (2). pj/ ptj
® a ^ Compiled by the sramana Sakya Tao-hsiian.’’ This
8
LIONEL GILES
is a A’inaya commentary, chapters 8-12, which will be found (together
with further commentary by ^ B8 Yiian-chao of th e Sung) in
K. Suppt. A. Ixix. 2 and 3. Of the 30 chapters, nos. 8-12 occur again
in S. 726. and others in S. 2328, 2525 (these have the j>reface and
table of contents). 4533. and 5404. It seems doubtful \=\'hether the
author can be the famous Tao-hsiian who, though living sut this time,
belonged to the 5 ^ Hsi-ming Monastery and died as late as 667
in his 72nd year : see Snuff Jmo sena chunv. ch. 14. no. 1.
Colophon :
"Preparation completed in the Ch'ung-i Temple at tie AVestern
Capital [Ch'ang-an] during the summer of the 9th year of A\’u-te of
the T ang " [626]. This is followed by another column of characters
which do not seem to yield any connected sense. BotTi title and
colophon are written in red ink.
The roll is about 59 feet long, and composed for the naost part of
a very thin fibrous paper, unstained. The last five sheets are coarser
and thicker than the rest. The gap of ten years between this roll
and the last seems to reflect the disturbances which acconapanied the
change of dynasty and which lasted until Li Shih-min was firmly in
the saddle ; even then, the output of sutra.s continues to be noticeably
smaller than before.
627 (TAng).
S. 2231. *Ta pan nieh p‘an ching, ch. 39 (corresponding: to K. viii.
6, ch. 39 and beginning of ch. 40).
This is a fine AIS. on thin yellow paper of the sixth cejitiiry. The
following colophon, added perhaps a century later, is careles-sly written
in another hand : —
t^m n — ^ M m m m 7%^ m A m m
AHflfflSt: [for %i\
*■ Ling-hu Kuang-ho, having in his possession an old tattered copy of
the Nirvana sutra. has had one section of it repaired, so that it may
be read and recited on behalf of all living beings, and that those who
hear the .sound of it may never fall into the three unhappy states of
existence or the eight calamities ; and he prays that they may behold
Amida Buddha. Repairs completed on the 8th day of the 2nd moon
of the 1st year of Cheng-kuan ” [28th February, 627J.
Ling-hu Kuang-ho is also recorded as the possessor of S. 539.
containing a section of the Chin kimng ming ching. For the eight
calamities, cf. 8. 39.35 (a.d, 583). The roll is over 28 feet
long.
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION
9
628 (T‘ang).
S. 1218. *Ta pan nieh p‘an ching, ch. 12 (N. 113, K. viii. 5, but
ending somewhat sooner than the modern text).
There is a note at the end in cursive script : ^
' Note made by So Ho-fa on the
22nd day of the 2nd moon of the 2nd year of Cheng-kuan " [1st April,
628]. This, like the preceding colophon, must have been written
many years after the sutra had been copied, for the latter is in
markedly archaic handwTiting which can only be assigned to the
first half of the sixth century. It is a fine bold MS. on thin dark yellow
paper, making a roll over Ilf feet long.
637 (T-ang).
S. 2838. ^ 0 fa M. * ching, ch. 3. p’in 10-14.
(N. 146.) This is our earliest dated specimen of the Vimalaklrtti-
nirdesa-sutra, which is one of the half-dozen most popular sutras
found in Tunhuang. It is a v'ery fine MS. on thin light brown paper,
smooth and compact, but rather brittle. The first sheet, made of
coarse yellow paper, contains the beginning of the chiian in an inferior
hand of much later date. The roll is nearly 32 feet long, discoloured
in parts, and patched here and there at the back.
It appears from the colophon, which is composed in a more
elaborate style than usual, that this copy was made to the order of a
daughter of a secondary concubine of ^ ^ Ch'ii Wen-fai, who
was at that time King of Kao-ch'ang. The country was conquered
by the Chinese three years later, when the dynasty came to an end.
See ^ Kao cJiang. by ^ ^ JS? Huang 4Ven-pi. ff. 16, 17.
Colophon :
'ft m. m m ^ .k ± ± ^ ^ m m m m ^ ^
m m ^ ±
-u m n m
H it
II i+c 111 Jjjf ^ 1; ^ t: X it IP] ^ il mm
m ± Mi ^ m m ^ m m m ^
"Copied by the .scribe Ling-hu
Shan-ku. Eevised bv the Master of the Law Ts‘ao Fa-hui.
10
LIONEL GILES—
Sanctioned by the Director of monastic diet in the Lotus
School, the fully ordained monk and sramana in charge of affairs.
Fa-huan. On the 3rd day of the 5th moon of ting-yu, the 14th
year of Yen-shou [1st June, 637], a female devotee (upasika)
makes obeisance and surrenders her life, to abide forever in the Tri-
ratna. Having heard of the flayed skin and the split bone, and liearing
in mind the dictum of the Mahasattva about losing the body and
sacrificing the self, she seeks the mvstic significance of Prajna. Now.
she knows that the true inspiration of the Holv Teaching, as set forth
in the Golden Compositions and the Jade Records [i.e. the precious
Buddhist scriptures], is hard to be grasped by eye oi l)y ear; yet.
since it is worthy of honour and veneration, this disciple, who ha.s
come to birth as the descendant of a princely liouse and has grown up
in the recesses of the palace, depending on the love and indulgence
of tlie King her father, and receiving instruction and admonition from
her mother, liis con.sort. .so as to be deeply im1)ued with the rich
influence of the Law, has on that account caused a copy to be made
of this sutra. with the intention of turning to it for guidance day by
day. reciting it morning and evening. And this small stock of merit
she begs re.spectfully to offer to her royal father, praying that his
sacred person may enjoy peace and happiness, and that all hi.s wishes
may be fulfilled. 8he prays, too. that her deceased ancestors who
have lived in the remote past, and all those connected with her by
blood and kinship, may .see the face of Buddha and hear his Law.
and be reborn in the Pure Land : that the remaining years of the
Dowager Con.sort [i.e. the King's mother] may be lengthened, and the
glory of the Royal Consort be augmented : that the Heir-Apjrarent
and the other princes may enjoy protracted longevity ; that thieves
and robbers may be driven away and dispersed, that disease and
pestilence may he abolished, that the people may feel the mercy of
kindly protection, and all men e.xperience the jovs of prosperity ;
and that all classes of beings po.s.scssing the breath of life and endowed
with consciousness and perception may alike be freed from the source.--
of suffering, and together rise to attain the wonderful fruit [of bodhi
and nirvana]."
Fa-huan appears again as directing or authorizing tin* tran-
scription of 8. 409 (N. 130). in exactly the same formula. 5^ 15
the last regnal period of the ^ Ch'ii dynasty at Kao-ch'ang
(a.d. 624-640). ^ : the story goes that in one of Sakya-
muni s previous incarnations a Brahman said to him : "I have here
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 11
a gatha containing the Holy Law. If you truly love the Law I will
give it to you.’’ He replied ; “ I truly love the Law.” The Brahman
said : “If you truly love the Law, you must use a piece of your skin
as paper, and write the gatha with your blood instead of ink. Then
I will give it to you. “ Sakyamuni immediately did what he was
told— breaking a bone, flaying his skin, and writing with his blood.
See ^ Ta chih iu lun. ch. 16, p‘in 27.
639 (T‘ang).
S. 3888. Redo: :1c ^ ^ *Tafa)iij teiip jit lai
tscnuj cJiing (N. 384).
Colophon : + H
0 111 sHt + H o ®!i ® .80 “ Dated the 10th
day of the 7th moon of the IGth year of Yen-shou [14th August.
639]. Copyist, Kimg Ta-tzu. 12 sheets of paper ii.sed. [Only three
remain.] Revised by T-an-hsien. Master of the Law."
Thi-s is another Kao-clrang roll, on similar paper. It is only 3 feet
long. Verso : Extracts from llVi >iio ch'i citing (N. 146), p’in 7, in an
indifferent, sprawling hand.
641 (TangI.
8. 4284. ~h ^ ^ ® $1 *Ta fang pirn fo pao en citing
(X. 431), ch. 7.
Colophon: ^ I'i ftli + Tt ^ -t: D A H 14^ fl5c ^
^ ^ ^ ^ is; m m m m S'] ic n ^
^ ^ t iit ia- ^ ® U - ;"li mm ^ ^
3C # ^ i ^ M [for fSJ I9fe [a f ; ^ ii
oil the 8th day of the 7th
moon of the 15th year of Cheng-kuan [19th August. 641]. the disciple
of Bodhisattva-pratimoksa Hsin Wen-hsiang. who lost his home and
became a waif, separated from his parents, and neither party knowing
whether the other were alive or dead, has reverentlv caused a section
of the Pao en citing to be copied on behalf of his gracious father and
his deceased mother, following it ivith a prayer that wheresoever his
parents mav be reborn they mav meet Buddha and hear the preach-
ing of his Law. and that thev may alway's be born in an honourable
station of life, without passing through the three unhappy states of
existence or the eight calamities. He also prays that his own future
parents and kinsfolk in this ivorld may never be parted from him.
that virtuous prayers may' be granted, and that all may come to a
condition of perfect enlightenment.’’
12
LIONEL GILES —
are the rules found in the sutra of that name (N. 1096),
taken from the Fan icang ching. For the eight calamities,
see SOS. VII, p. 831.
This is a good bold MS. on crisp golden-yellow paper, somewhat
different in texture from that of the Sui. The roll is over 15|^ feet long.
652 (T-ang).
S. 3394. *ire{ mo ch‘i ching, ch. 2, pin 7-9. There is no colophon
beyond the following note : ^ ^ H ^ -f* 3£ 0 ft)
^ 7C ^ ^ ’ Acquired bj" the Buddhist disciple Teng Yiian
on the 15th day of the 5th moon of the 3rd year of Yung-hui ” [26th
June, 652].
This is a very good. weli-.spaced MS. on rather soft yellow paper
of excellent quality. The roll is about 19J feet long.
669 (T-ang).
S. 5181. ft hsing hai tsang ching, ch. 1. This
is an apocryphal sutra on " the accumulation of knowledge and
wisdom", tile full title ft ^ ^ ^ ^ M <£!' II
being given in S. 2169, where the whole of chiian 1 is preserved.
There are also two copies of chiian 2, contained in S. 4000, S. 4103.
Note at the end : ^ GH A ^ -fr $ 0 ft ^ f‘Jl ^
^ ^ M. The 25th day of the 8th moon of the 4th year of
Hsien-ching [16th September, 659]. The Buddhist di.sciple of pure
faith [upasaka] So Huai-tao.’' This person may be the owner or copyist
or both.
The roll is over 23 feet long, and made of paper of very good quality,
stained a rich yellow. The handwTiting is less elegant than in the
preceding roll, but large and clear.
663 (T‘ang).
S. 4656. ig fj; *I chino ching. This is N. 122, a sutra spoken
by Buddha just before entering Parinirvana, with abbreviated title.
The colophon is written in the .same hand as the text of the
sutra :
M m m m ~ ^ m ^ m fi m. m Wl m 'j^' m ^
^ is m ^ JE tn ± m ^ ^ ^ m ^
M i m m '^7 m ^ i ^ if ^ n aE m Mr
m ^ m it M ^ m ^ m m m — ^ n it
m. m • n : fE' it ^ ^ Ei m ^ % it u —
m m R a # # ^ ^ "p ff m nR m
A ^ P'l I?] ^ ^ [ft it o SI ^ + v:
H -H- H 0 frli.
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 13
Here it is recorded that the sutra was copied to the order of the
true believer Liu Ching-an and his family, and completed on the 23rd
of the 12th moon of the 2Ed year of Lung-shuo [6th February, 663].
As several passages, including an elaborate exordium on the subject
of the Dharmakaya. are decidedly obscure, I shall leave the translation
to those more deeply versed in Buddhist terminology. One striking
piece of h\^erbole, however, may be noticed ; a single recitation, it
is said, of the speech that i.s.sued from the Buddha'.s '' golden mouth
surpasses in value the sacrifice of as many lives as there are grains of
sand in the Eiver Ganges !
The roll, made of thin buff-coloured paper, is nearly 6 feef long.
The handwriting is only moderately good.
668 (T ang).
S. 4496. *Miao fa lien hua ching, ch. 3, p'in 5-7.
Colophon: — H-fj-HElfftlfe ^-1#
''s ^ n ^ xt % n ^ X: % n ^ m m
)§, -fit ^ On the 23rd day of the 11th moon of the 1st year of Tsung-
chang [31st December. 668] the Buddhist disciple Yin Chih-chu. on
behalf of his parents now living and his parents of previous incarnations,
reverently caused a section of the Fa hua ching to be copied for circula-
tion, as an act of worship."
A good MS. on a rich golden-yellow paper of the best quality.
Length of roll, 30J feet ; width, 25-5 cm.
670 (T‘ang).
S. 3655. *Miao fa lien hua ching, ch. 7. This is only a fragment,
26 by 24 cm., from the end of p in 24.
Colophon: B Kt
ia Reverently made to the order of Tsnii An-chii on behalf
of his deceased elder sister. Madame Chung, in the intercalary 9th
[moon] of the 1st year of Hsien-heng [20th Oct. -17th Nov.. 670]."
671 (T‘ang).
S. 2215. *Miao fa lien hua ching. ch. 2. p in 3. 4.
Colophon: ^ ^ %
t ^ @ Lr S I’l^I & ^ " t'opy reverently made to the order
of the disciple Fan Huai-hsin on behalf of his deceased wife Chao and
his deceased nephew [brother’s son] A-nu. in the 12th moon of the
1st year of Hsien-heng [17th Jan. -14th Feb.. 671]."
At the end of this colophon a note has been scrawled in verv faint
ink : Jt ?/, flL 5 " Received by Tung Hung-chi. " This is a neat,
well-spaced MS. on yellow paper.
14
LIONEL GILES —
671 (T‘ang).
S. 5319, 84, 3079. *Mi<io fn lien hna ching. ch. 3. p'in 5-7 ; ch. 5,
p'in 15-17 ; ch. 4, p’in 8-13. These three rolls form part of a series
in which the colophon is cast into tabular form. That of 8. 5319 runs
as follows : —
^ ^ S -ti- - FI # ¥ ^ it
m m m -V X m ^ m m ^ m
m m X m m m m m x m m m
%f m xm i m # ga y. ^ it
m m X m m m m m ^ m m f,
X ^ 3L m X ^ ±i^#_h)iiikbti^
m. ^ m fU ^ m k iS W s-f if it (5 ft
k m w if ff ^ z ft
■P\ 'k iff gt ^ m 62
iSi ± ^ k k iy 'k ^ ^ m n 'p E
7ic n iM rM
•• Copied by the scribe Ch'eny Tu on the 22nd of the 5th moon of the
2nd year of Hsien-heng f3rd July, 671]. 19 sheets of hemp paper used.
[18 remain.] Dyer of sutras. Wang Rung. Carefully perused by the
Ijhadantas Ling-pien. Chia-shang. Hsiian-tse. Ch’ih-shih. Po-ch‘eu.
and Te-lisiui. .Superintended by Hui-li. the active director, and Tao-
ch'eng, the spiritual director (inahasthavira) of the T’ai-yiian Monastery.
First revision bv the copyist Ch’eng Tu ; secoiul revision by Ta-tao, and
tliird revi.sioii by Chih-an, both priests of the Ta-tsung-ch'ih Monastery.
Hsiang I-kan. p'nn knun. etc. Under the general superintendence of
Yii Ch'ang fa high official at Yung-hsing Hsieii. i.'V Ch'ang-an], etc.”
Contrary to what one miglit e.xpect. tli<‘ actual co])viiig of sfitias seems
to havidx'ondone more often l>y a lavmanthan bva monk. Here Cffi'eng
Tu is entitled ^ ^ " scribi' ” as well as 0- p ■' sutra cojiyist ”.
and also dignified by tlie unusual addition of ^ t<j liis name. For
^ under the date (>12. abov(>. Hui-b is the well-known
priest who compiled the life of H.siian-tsang (X. 1494) si.v vears before
the date of this M8. His name appears in nianv otlnu’ colophons of
t!ii> Collection. In Snug l(ir> yetig rlninn. ch. 17. no. .3. it is given as
X- but as it is also stateii that he became aeti\'e (lir“ctor of the
T'ai-yiian ilona'-tery. there can be no doubt as to his identitv. Accounts
of Chia-shang and Tao-clreng will be found in the same work. ch. 4.
no. 9. and ch. 14. no. 2. respectively. Two stroki's are oniitti'd in the
character ]Jb for reasons of taboo. lyr K 1 icing the per.sonal name of
the second T’ang emperor. 8. 84 and 8. 3079 are dated the 16th and
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 15
ISth November, 671. respectively. In both, the copyist and first
reviser is f|5 Kuo Te. the dyer is ^ ^ Hsieh Shan-chi. the
second reviser is ig Fa-hsien. and the bhadanta ^ ^ Shen-fu
is one of the readers. In S. 84. ^ ^ P'u-ting appears as third reviser,
in S. .3079 ® (for Ssu-k'an. All the rest are the .same as in
8. 5319.
These are all finely written MSS. on thin, crisp, dark yellow paper,
exceedingly hard and smooth, but with a tendency to brittleness.
672 (T’ang).
S. 4209, 4551. *Miao fa lien laia ching, ch. 3. p'in 5-7 : ch. 4.
p'in 8-13. These two rolls continue the series with tabulated colophons.
The former is dated tlie 17th May. 672. the copyist and first reviser
being ^ Chao M’en-shen, the .second reviser 'I',' ^ Chih-
tsang, and the third reviser Chili-hsing. Tlie latter is dated
the 2.jth September. 672, the copyist and first reviser l)eing §|J
Liu Ta-tz ii. tlie second reviser j|iL Hsing-li. the tliird reviser itji
Hui-ch'ung. The other names are as found in S. 84. S. 4209 is said
to be composed of 19 sheets of /J> “ little hemp paper ; but
the material appears to bo exactly tlie .same as that of S. .5319 and the
rest of the series.
672 (T-ang).
S. 36. H'] l|^ ^ If ^ IS '*Ch)ii huuj pan jo po lo
nii ching (X. 10). Anotlier roll with tabulated colophon similar to tlie
preceding. The date is the 19th June. 672. and the copyist. ^ ft
AVu Yiian-li. is called fE ^ ijj ^ ^ " writer in the clerkly style
attached to the left division of the Crown Princeis staff”. The three
revi-sions were all done by ^ )ji^ Hsiao I. The roll is 134 feet long.
673 (T’ang).
8. 2573, 312. *Miao fa Urn haa dung. ch. 2, p in 3-4; ch. 4.
p'in 11-13. A continuation of the .series of the Lotus 8ritra. The
two colophons are exactly the .same, excejit that the former is dated
1st November and the latter 5th November, and that in 8. 2573,
through an oversight, the name of Hui-li is omitted, while that of
Tac-ch'eng is repeated. ^ ^ h Feng An-ch'ang is the copvist.
the name of the dyer is given as M Hsieh Chi (cf. 8. 84 and 8. 3079).
the hist revision was done by “fg ^ Huai-fu of the Ta-
chuang-yen Monastery, tlie second and third by ^ Hsuan-chen
of the 0 HI) Hsi-ming Alonasterv. and the p'an-kaan is now ^
Li Te. Air. Clapperton describes the paper of 8. 312 as follows : "A
dark buff paper of very even texture. Thickness -004 inch. This sheet
16
LIONEL GILES
looks like a wove paper, as there is no sign of laid lines when looking
through the sheet. The laid lines can be seen indistinctly by reflected
light. Very smooth surface, without hairs. A short-fibred hard paper,
the fibres being shorter than in any other paper e.xamined up to thi.s
date, and the look-through of the paper is also very much closer.
Very evenly felted, and might easily pass for a close-wove paper.
Composition ; Paper mulberry."
674 (T'ang).
S. 456, 3348. *Mi(iofu lien hm citing, ch. 3, p in 7 ; ch. 6. p in 22.
23. Two companion rolls, both somewhat mutilated, dated the 2nd of
the 8th moon of the 5th year of Hsien-heng [7th Sept.] and the 25th
of the 9th moon of the 1st year of Shang-yiian [29th Oct.], respectively.
The seeming discrepancy is explained by the fact that the nicn-hao
was changed to Shang-yiian in the 8th moon. In S. 3348, half the
colophon is missing. Hsiao Ching is the copyist. ^ ^ Chih-
yen of the |g Fu-lin Monastery the first reviser. Iff ^
Hsing-kuei and ^ Huai-tsan. of the Hsi-ming Monastery, the
second and third. The p‘an kuan ^ ^ Li Shan-te is evidently
the same person as Li Te in the preceding rolls. Compare the case of
Hsieh Chi, above, who also figures as Hsieh Shan-chi.
675 (T'ang).
S. 1515. % m. 9 M & *Wu Hang shou kuan ching. This
is the Fo shuo wu Hang shoufo ching of N. 198, K. x. 4. i. The roll in
its present state begins at the end of the 7th of the 16 fg meditations.
Colophon
m m -k m R m ‘t' m ^ m 9 m m - u R m
w m - m B lib Jjj m ± n r m k is m
^ m r R A: R x n H R m ^ M ^ w. m m
P*] fft ^ ^ i " the 28th day of the 4th moon of the
2nd year of Shang-viian in the Great T'ang dynasty [28th May, 675]
the Buddhist disciple and upasika the Lady Chang, with pious intent,
has reverently caused copies to be made of the Wu Hang shou kuan
ching and tlie Kuan gin ching, praying that this act of merit may,
fir.stly, benefit our Divine Emperor and Emprc.ss. to the indefinite
prolongation of their sovereign influence, and secondly, affect her
parents of seven previous incarnations and all the living creatures of
the \inivcr.se, .so tliat they may escape from the gates of affliction
and one and all a.scend to the wondrous realms of puritv."
The roll is 14f feet long. Apart from other considerations, the
heavily oiled yellow paper and fine handwriting would make it fairly
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 17
certain that the Shang-yiian period here mentioned is not the later
one of 760-1. The Kuan yin ching is ch. 25 of the Lotus Sutra, often
regarded as a separate work. The emperor in question is Kao Tsung,
and the empress is the notorious ^ ^ij 3 ^ Wu Tse-t‘ien, who subse-
quently usurped the throne for over 20 years.
676 (T‘ang).
S. 114. *Miaofa lien hua ching, ch. 7, p‘in 24-28.
Colophon :
^ “ Reverently copied to the order of the upasaka Chang Chiin-
ch‘e on behalf of his deceased younger sister in the 3rd year of Shang-
yiian.”
^ may simply be a title of respect, as in the case of Ch'eng Tu ;
see S. 5319 (a.d. 671). The name of the upasaka would then be Chang
Ch'e. For all we know, the “ deceased younger sister ” may be no
other than “ the Lady Chang ” who was herself causing sutras to be
copied in the previous year.
This is a fine MS., rather exuberant in style, on yellow paper of
excellent quality.
676 (T‘ang).
S. 2181, 1456, 3361, 2637, 4168, 1048. *Miaofa Ueti hua ching,
ch. 2, p'in 3, 4 ; ch. 5, p‘in 14-17 ; ch. 1, p‘in 1, 2 ; ch. 3, p‘in 7 ;
ch. 3, p‘in 5-7 ; ch. 5, p‘in 14-17.
I have arranged this series of rolls according to date. All are
imperfect or badly mutilated at the beginning. The scribes are laymen
and all different, but the paper-dyer is the same in each case, namely
Hsieh [Shan-] chi : see years 671 (S. 84) and 673. Li Te is still the
p‘an-kuan, but the general superintendent is now ^ ^ jS Yen
Hsiian-tao. S. 2181 is a fine MS. dated the 1st June, but the greater
portion of p‘in 3 has been added in a different hand. S. 1456 is dated
the 29th June ; the monk who is responsible for the three revisions
of the text bears the curious name of ||l Fa-chieh (Dharniadatu,
the universe, things in general, or their underlying cause) and hails
from the itM Hua-tu Jlonastery ; this was a temple of the Three
Stages sect founded hy ff Hsin-hsing. S. 3361 is dated the
11th September, and S. 2637 the 13th September. The copyist of
the latter text, Jen Tao, might be taken for a monk, but that
he was a member of the ^ Hung-wen Kuan, an official
College of Literature, 'ff, however, is a fairly' common surname. The
first reviser, m R w u-chi, belonged to the ^ Tz'u-men
Ssu, which was also a temple of the Three Stages sect. S. 4168 is dated
VOL. VIII. PAKT I.
18
LIONEL GILES
the 20th October. For S. 1048, dated the 15th December, 21 sheets
of “ little hemp ” were used. It is a thin, crisp, brownish-yellow paper
of even better quality than usual at this period.
676 (T‘ang).
S. 513. *Chin hang pan jo po lo mi ching. This, though four years
later in date (29th April, 676), may be regarded as a companion roll
to S. 36. The copyist is ^ @ Ou-yang Hsiian-che.
677 (T'ang).
S. 4353, 2956, 3094. *Miao Ja lien hua ching, ch. 1, p‘in 1, 2;
ch. 7, p‘in 27, 28 ; ch. 2, p‘in 3, 4.
These are three fine MSS. on the famdiar brownish-yellow paper.
S. 4353 was copied on the 2nd January, and S. 2956 on the 29th of
the same month, both by ^ ^ ^ Wang Chih-wan of the Hung-
wen Kuan. The revisers and readers, too, were the same in each case.
S. 3094 was copied on the 26th June by fl] ® Liu I-shih, and
revised three times by :fij Liu Yen.
678 (T=ang).
S. 3135. -jiz ^ ^ W: *T‘ai hsiian chen i pen chi
ching, ch. 2. There are four other rolls in the Stein Collection contain-
ing parts of this Taoist sutra, which does not appear to be included in
the present Canon. Taoism was in the ascendant during the T'ang
djTiasty, the rulers of which claimed to be descended from Lao Tzu,
though it suffered a little set-back under the more Buddhistically
inclined Empress Wu. The colophon is an amusing example of the
slavishness with which the devotees of this bogus religion aped the
ideas and phraseology of their rivals, the Buddhists : H ^
H D -H- - 0 H ^ ^ ^ ^ t
“ On the 22nd of the 3rd moon of the 3rd year of I-feng [18th April, 678]
the female official of the Three Profundities, Kuo Chin-chi, reverently
caused a section of the Pen chi ching to be copied on behalf of her
deceased Preceptor, in order that he might be helped and benefited
by the resultant stock of surpassing happiness, praying that his path
might coincide with that of the Nine Hsien (' Immortals '), and that
his spirit might travel to the Eight Bles.sed Eegions.”
The works constituting the Taoist Canon fall into three main
diffisions which are called “ Grottoes " or “ Profundities ". corre-
sponding to the three Pitaka of the Buddhists. They are : (1)
^ the section of Profound Purity; (2) ^ ^ the section of
Profound Mystery ; (3) ^ ^ the .section of Profound Spirituality.
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 19
The first is supposed to embody the teaching of tc ^ ^ the
Primordial Heaven-honoured One, the other two that oi ^ ^ ^
the Most Exalted Lao Chiin (Lao Tzu).
It is interesting to find a group of Nine Immortals mentioned
instead of the familiar Eight. The cult of the Nine appears to have
been widespread, for we find the following passage in the ^ !ltl 0
Teng choufu cJiih, ch. 18, f. 20 ; “ Sung dynasty : the Nine Immortals
with one eye between them. One of the Nine had the sight of one eye,
the other eight were blind. They got the one-eyed man to lead them,
and they begged in the streets as they went. At night they halted under
^ Guests’ Bridge in the north-west of the city [of Teng-chou.
Shantung]. No one knew who they were. After a time, the prefect
of the city met them at the bridge and issued a proclamation concern-
ing them. But suddenly the Nine underwent transformation and
vanished. Wherefore the name of the bridge was changed to jf]] {ill
Ying-hsien (i.e. ‘ AVelcoming the Hsien ’).” There is also a long account
in. # fllj IS Shen hsien rung chien, vii, sec. 7-8, of magical
feats which they performed in the presence of the first Han emperor.
In this work they are said to have been brothers, belonging to the
Ho family. A mountain near the city of Foochow is still called {ill lU •
The /\ ^ form a counterpart to the Buddhist Pure Land. The
number is doubtless suggested by /V the eight points of the compass.
After the colophon, there is a note in red ink '■ [3 ti “ [ ] recorded.’’
This is a fine MS. on rich yellow paper. The roll is about 14 feet long.
684 (T‘ang : usurpation of the Empress AVu).
S. 286.3. [*Miao fa lien hua ching, p’in 25.] The beginning is
mutilated, but only slightly imperfect. There is no title at the end.
Colophon : X m ^ ^ n 31 B m ^ m m n
Copied and recorded by the
disciple So Jen-chieh on the 5th of the 6th moon of the 1st year of
AV en-ming [22nd July, 684]. He prays that his parents of seven
previous incarnations, as well as the father and mother who begot
him [in the present life] may be reborn in the realm of Amitabha
Buddha in the AA'estern Regions, and that [the merit acquired] may
also be shared by his brothers and sisters. Such is the virtuous prayer
he would constantly utter.”
^ 0JJ lasted from the 2nd to the 9th moon of this year. It was the
first nien-hao taken by the usurping empress, and as such is ignored
by official historians, who continue to use Chung Tsung's year-title
20
LIONEL GILES —
M H until his restoration in a.d. 705. The colophon is written in
the same uncultivated hand as the text of the sutra, which does not
suggest a professional cop}'ist. The paper is coarse and undyed, but
fairly tough. The roll is about 8| feet long, and 27-5 cm. wide.
688 (T'ang ; usurpation of the Empress Wu).
S. 194. *Miao fa lien kua ching, ch. 1, p‘in 1, 2.
Colophon
^ " On a day in the 6th moon of the 4th year of Ch’ui-kung
[3rd July -1st August, 688] the female believer Yang A-seng gave
[this roll] as a lasting possession to her fellows."
A good bold MS. on yellow paper, 12 feet long.
688 (T'ang ; usurpation of the Empress Wu).
S. 791. *Miaofa lien kua eking, ch. 3, p‘in 5-7.
Colophon :
^ it M ^ -^m m ^ nm m m ~
K S (ft "In the 11th moon of the 4th year of Ch‘ui-
kung [29th Yov.-27th Dec., 688] the wife of Wang Lin, a Buddhist
disciple of pure faith, being afflicted with a chronic disease, has
reverently caused a section of the Fa hua ching to be copied, in order
that this act of merit may benefit all beings endowed with consciousness,
and that all may attain the fruits of Buddhahood.”
A good MS. on rather soft yellow paper, about 28-^ feet long.
689 (T'.a.ng ; usurpation of the Empress 5Vu).
S. 592. *Miao fa lien kua eking, ch. 2, p‘in 3 (slightly imperfect). 4.
Chlophon:
m ~ iif, m t: m R M
^ ^ ilt •• Iuthel2thmoonofthe4thvearof Ch‘ui-
kung [28th Dec.-26th Jan., 689] Madame Chb, the wife of Wang Lin
and a Buddhist disciple of pure faith, has on behalf of her deceased
daughter reverently caused a .section of the Fa liiia to l)c copied,
praying that the deceased, as well as all living beings in the whole
universe, may together achieve Buddhahood.”
This roll was made a month later than the one preceding and also
to the order of Madame Wang, who.se maiden name we now learn
to have been of which $ is a vulgar form. The roll is 32i feet long,
and made of excellent paper dyed a bright yellow.
691 (Chou : usmpation of the Empress IVu).
S. 2157. *Miaofa lien kua eking, ch. 4 (end only)
M„ph„„ (s^e mm ittammm » 0
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 21
■‘ The bhikshunl Shan-hsin of the Ling-hsiu Nunnery, aware of the
unreality of the body, the shallowness of perception, and the illusory
nature of suffering, has therefore, on behalf of her deceased mother,
cut down her personal effects over and above the [necessary] three
garments, and reverently caused a section of the Fa hua ching to be
copied, praying that through the merit of this virtuous act her mother
may avail herself of the resultant stock of happiness and be reborn in
the Pure Land, in the highest grade of the highest class, and may
at the present moment enjoy peace and comfort ; also, that all sentient
beings may participate therein, and together taste the wonderful fruit
[of bodhi and nirvana]. Copied on the 29th day of the 3rd moon of
the 2nd year of T’ien-shou [2nd May, 691],”
The Ling-hsiu Nunnery is mentioned several times in the Stein
MSS, From a list of monasteries and nunneries given in S, 2614 v°,
which may date from the tenth century, we learn that it then contained
a total of 142 inmates, comprising 99 jg nuns who had taken
the full vows, 29 ^ ^ Jg probationers (siksamana), and 14 ^ /g
novices (sramanerika), of whom two were senior and twelve junior.
H 56c are the three regulation garments ^ kasaya) worn by
monks and nuns, consisting of a vest or shirt, an upper garment, and
a “ patch-robe ”, reaching from the shoulders to the knees and fastened
round the waist. — There are three classes of sentient existence in the
Pure Land, upper, middle, and lower, each similiarly divided into
three grades, making nine in all. — This is the earliest dated MS. in
which we find the new characters adopted in 689 by the Empress M'u
on the recommendation of a minister called ^ M ^ Tsung Ch'in-
k‘o. As will be seen by reference to Plate I, the characters in the
above colophon which appear in their altered form are
^ , and 0 • Of these, ^ does not occur in the list of 12 characters
given in T‘ang shu, Ixxvi, 10 v°, and repeated with a little variation
in Tzu chih Vung chien, cciv, 19 v°, and Ka»g mu, xli, 106-7. On the
other hand, it does occur in the list of nineteen new characters given
by fn m Hsiian ho shu p‘u (reproduced in shu chi ch'eug,
xxiv, 1, f. 9 r°), and also in the list of sixteen given by Cheng Ch'iao
in his T‘utig chih. ch. 35, f. 13 r°. The latter points out, however,
that the altered form is not really new, but derived from the archaic
script. This MS. is a small fragment only about 1 foot long. The
handwriting is very clear and good.
90
LIOXEL GILES
692 (Chov ; usurpation of the Empress u).
8. -238. ^ it 3i ^ A ^ ^ ^ *Chin chen ijii Jcmui'
pa cilia g fei ching. This is another Taoist sutra which is no longer ti
be found in the Taoist Canon. It is devoted principally to the descrip-
tion of various deities, and blank spaces have been left in the text fn:
their portraits.
Colophon : 7C ^ S id “t* ^ 0 ^
m m m ^ Mm m m m m m m ~
■■ Copied by the scribe dVu Chung on the 13th of the 5th intercalar\
moon of the 1st vear of Ju-i [2nd July. 092] to the order of Chih-sU'
of the Ch‘ing-tu llouastery. with his colleagues Ssu-chieh, Chu-yung
and Clii-chben. 18 .sheets of paper u.sed.” [10 of these remain.]
This colophon well .shows the di.stinction in meaning between
and ig. The names of the Taoist monks are rather strange, e.g. Chi-
ch'ien means ” Slum-money ". Ju-i lasted from the 4th to the 8th
moon of this year, inclusive. The Empress Wu characters in the
colophon are and 0. In the text of the sutra. the ordinary
forms of all three occur .several times.
A fine ILS. on bright yellow paper of particularly good quality.
The roll is 16C feet long.
694 (Cnor : usurpation of the Empress Wu).
S. 3542. 'fiij', Ig ijiJ; pg *Fo skuo a an t'o ching. This h
the first dated example of the so-called " smaller Sukhavatl-vyfiha
in Kumarajiva's translation (N. 200). which always remained far more
popular than the later and fuller translation by Hsiian-t.sang (N. 199) '•
of the latter there is but a single copy in the Stein Collection a.s opposed
to several dozen of the other.
Colophon: H
3ia M 'jT' K — ■ Rp '■ On tlie Lst day of the 6th moon of the
3rd year of ClCang-shou [28th June. 694] the Buddhist disciple Master
Ti reverently caused a copy to be made of the Arnita Sutra.”
Three of the Einpre.ss M u characters occur here : ^ , and 0 •
A good MS. on soft yellow paper. Boll about 5i feet long.
694 (Chou : usurpation of the Empress Wu).
S. 5176. *MiaoJ(i lien him ching, ch. 3, p‘in 5-7.
Colophon: -k m ^ ^
^ ^ n t: m m m n ~ U ‘^On the Ith day of the
4th moon of the .frd year of Ch'ang-shou in the Great Chou dynasty
[6th May, 694] Seng Huai-chen of the Ta-yiin [Great Cloud]
Bl LI.. S.O.S. VIII. 1*.\KT 1.
I’l .M K. I
[ 4 ;
/ ■' r-" ‘ p j
l^oUR Colophons srowin<; -Empukss Wo Ch.xhaotkrs ' . [..\tki,
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 23
Monastery reverently caused a section of the Fa hua cliing to be
copied on behalf of his deceased mother.”
The name of the djTiasty had been changed from T‘ang to Chou
by order of the Empress in 690. The same three new characters occur
again in the colophon.
A very good, well-spaced MS. on rich yellow paper. Eoll about
30^ feet long.
695 (Chou : usurpation of the Empress Wu).
S. 2278. {i)h tfe ^ IS *Fo shuo pao yii ching (Ratna-varsa-
stitra), ch. 9. This is the only copy of N. 151 in the Collection. It
contains the longest and most elaborate of the tabulated colophons,
beginning thus;
Translated at the Fo-shou-
chi Monastery on the 3rd day, chi-ch‘ou, of the 9th moon, the first
day of which was ting-hai, of the 2nd year, hiei-ssu, of Ch‘ang-shou
in the Great Chou dynasty [7th October, 693].”
The characters if, ^ , 0 > and ^ are written in the new style.
The Fo-shou-chi (Buddha-given Record) Monastery stood inside the
^ Chien-ch‘un Gate at the capital. Its original name ^ ^
Ching-ai was changed by Huai-i after the buildings had been extended.
The colophon continues :
^ ^ “■ The bhadanta and sramaria Huai-i of the Ta-pai-ma (Great
White Horse) Monastery supervised the translation.”
Huai-i was a favourite of the Empress Wu, and a notorious
scoundrel. He was a native of ^ 0; Hu Hsien near Ch‘ang-an, and
his original name was /J'. ^ Feng Hsiao-pao. Noted for his size
and physical strength, he was taken by the Princess Ch‘ien-chin
{=f- ^ ^ ^) into the Palace, where he soon became the secret
lover of the Empress. In order to divert suspicion from his comings
and goings, she had him ordained as a Buddhist priest and made him
Director of the "White Horse Temple. She also forced ^ ^ Hsieh
Shao, son-in-law of the Princess T‘ai-p‘ing (;;;fj zf! ^), to adopt
him into his clan, and he was thenceforth popularly known as m ®
the Preceptor Hsieh. Huai-i now gave himself up to all kinds of
lawlessness and debauchery, and a brave censor who ventured to
impeach him was waylaid and heateu almost to death. He was
appointed to superintend the reconstruction of the ^ Ming T‘ang,
and his services were rewarded by the titles of ^ ^ ^ ^ ^
Commander-in-chief of the Left division of the Imperial Guard, and
^ S Duke of Liang. After a military expedition against the
2i
LIONEL GILES
T'u-cMeh had added to his honours, he was employed with a number
of other monks in forging a so-called ^ ^ ^ Great Cloud Sutra,
in which the Empress was extolled as an incarnation of Maitreya.
But seeing his influence begin to wane with the rise of a new favourite,
in a fit of jealous anger he set fire to the Ming T‘ang and the Temple
of Heaven, also newly erected, and both were burnt to the ground.
Though his share in this crime was hushed up, his ever-increasing
arrogance alienated the Empress, and the discovery of a treasonable
plot finally led to his being seized and strangled. This happened at
the end of 694, only six months before the date of this MS. The above
account is taken from Chiu t‘ang sku, clxxxiii, 15-16.
Colophon (continued) :
% 4: O PP I ^ ^ ^ ^ 4^ “ Sanskrit
text published and explained by the sramana Dharmaruchi of Southern
India ; Sanskrit text conjointly published by Fan-mo, royal envoy
and sramana of Central India.”
By order of the Empress, Dharmaruchi’s name was afterwards
changed to ^ gft * Bodhiruchi, and it is under this name that
he appears in Sung kao seng chuan, ch. 3, no. 4.
Next come four monks who played an active part in the translation,
one as interpreter ®), one as check-interpreter ^ ^),
and two as verifiers of the Sanskrit (f^ % *)• Here we find ^
written in the new style.
Two monks appear to have taken down the words in actual dicta-
tion (U ^), two igg “ sewed together ” or linked up the composi-
tion, and no fewer than nine ^ ^ verified the meaning. A Brahman
priest verified the translation, and a Brahman minister with a Chinese
name, ^ ^ gf} Li \Vu-ch‘an, acted as interpreter. Here g is written
in the new style. Three Brahmans and a native of ^ llj Ch‘ing-
shan Hsien in ^S| ^ Hung-chou (Shensi) copied the Sanskrit text.
The minister ^ ^ ^ Li Shen-kung, shatig-fang in charge of the
craftsmen (f^ ^ ^ g), prepared the materials (^), i.e. paper,
ink, and dye. Shang-fang was the name of an officer concerned with
articles destined for imperial use. The list concludes with the names
of two specialists on manuscripts and composition, and two special
commissioners.
Then follows what may be regarded as the real colophon, written
in a much larger hand (see Plate I):
m n ^ m a h z. m ^ m
^ ft ^ M Wi ^ “ Collated and copied by the monk Tao-li,
DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION 25
cognizant of merit, on the 8th day, i-yu, of the 4th moon, the 1st
day of which was inou-yin, of the 1st year, kuei-xcei, of Cheng-sheng
[26th May, 695]. Examined and revised by his friend the monk Fa-lin.”
Cheng-sheng fell in ^ i-wei, not liuei-wei, and did not last the
full year. It is strange to find such a mistake in this elaborate colophon.
The Empress Wu characters are gg, if, ^ . and 0 . At the
very end of the roll, which is about 22 feet long, are the characters
-f* 7 ^ “ 16 ”, probably referring to the number of sheets, of which
only fourteen, however, remain. Paper and handwriting are both
very good.
695 (Chou : usurpation of the Empress Wu).
S. 5005. SP ^ *I’ao shih ching. This is N. 171, K. ix. 9. f,
with an abbreviated title.
Colophon; :k ^ M M ^ ^ H “hAH
^ ^ fi[) ^ ^ ■■ Reverently copied by the Buddhist
disciple of pure faith Lang Hsing-hsin on the 18th day of the 4th moon
of the 1st year of Cheng-sheng in the Great Chou dynasty [5th June,
695].”
The Empress Wu characters are the same as in the preceding roll.
A good MS. on good yellow paper. Roll 15 feet long.
696 (Chou : usurpation of the Empress M'u).
S. 217. -tfr # *Kiian shih yin ching. This is another
name for Miaofa lien hua ching, pTn 25.
Colophon (see Plate I):
R a 0^ R ^ ^ n m ^ u — ^
“ On the 15th day of the 1st moon of the 2nd year of T‘ien-ts‘e-wan-
sui [23rd February, 696], the Buddhist disciple of pure faith Yin
Ssu copied out the Kuan shih yin ching in one roll on behalf of his
parents now living and his parents of seven previous incarnations,
as well as on his own behalf and that of the multitude of living beings
in the universe." The Empress AVu characters are if, jE,
and g .
A good MS. on dull buff paper partly dyed yellow. Roll about
4| feet long.
700 (Chou : usurpation of the Empress Wu).
S. 87. *Chin hang pan jo po lo mi ching (N. 10).
The colophon (see Plate I) has a more worldly flavour about
it than usual :
26 DATED CHINESE MANUSCRIPTS IN THE STEIN COLLECTION
“ On the 23rd day of the 5th moon of the 3rd year of Sheng-li
[14th June, 700] Yin Jen-hsieh, Assistant Commissioner at Ta-sheng-
pa-ku, shang-chu-Jcuo, and Jc‘ai-kuo-ku»g of Nan-yang Hsien, caused
this sutra to he copied on behalf of the Holy and Dmne Sovereign
Lord of the Golden 4Yheel, his parents of seven previous incarnations,
and all the members of his family, great and small. He made a vow
that if he was promoted to the sixth official grade he would have one
roll copied every month, and that if he was promoted to the fifth grade
he would have two rolls of a sutra copied every month. But for a
long time, owing to warlike operations, paper and ink have not been
procurable, so that he did not fulfil his vow. Now at last, materials
having been procured, he has been able to have this copy made, to
be unrolled and read on behalf of all without exception.”
Nan-yang Hsien is in Honan, but the name Ta-sheng-pa-ku suggests
a place outside the borders of China proper. The cakra, wheel or
disc, is an emblem of sovereignty, and the ^ I or Golden Wheel
King is the highest of the cakravartin, a conqueror of the universe.
The Empress Wu, on whose behalf the sutra was copied, had assumed
the e%'en higher title in the text. Her special characters in this colophon
are 0. and Note the different form of which
also occurs in other 3ISS.
This is a very good IIS. on bright yellow paper, but the first sheet
has been added in a different hand. The roll is about 15 feet long.
{To he coniinuetL)
\
The Japanese Particles Wa, Ga, and Mo
By S. Yoshitake
OINCE Motoori Norinaga invented the term I'akari. lit. “ nexus ”,
^ more than 150 years ago Japanese scholars have been labouring
to find out the precise signification of the word he had left unexplained
beyond vaguely applying it to several grammatical devices which
included the uses of the particles wa, mo, zo, and Jcoso. Among those
who made a special investigation of this perplexing term is Yamada.
who has written a number of valuable books on the language and
literature of Japan. After many years’ study he came to the conclusion
that by the word hihari Jlotoori must have meant that the word or
group of words immediately preceding any one of the aforesaid
particles calls for a certain statement which he named muvubi, lit.
“ conclusion ” (YY., p. 52).
Based on this interpretation Yamada gives a long discourse, of
which the following is an outline. When anyone says Tori wn tobu toki
we feel that something is lacking, whilst no such feeling creeps into
our mind when we hear Tori ga toba toki. This is because ga connects
tori “■ birds ” with tobu fly ”, and there the matter ends : the particle
has no influence upon the element which is to follow toki " when
In other words ga indicates a subject and so long as it is joined to tobu,
the companion of tori, its duty is performed. But Tori wa tobu toki
requires some explanation as to what birds do or what happens to them
when they fly. Thus in the expression Tori wa tobu toki (ni) hane u’O
konna fu ni sum “ Birds, when they fly, move their wings like this ”
it is clear that tori wa has no direct relation to tobu but introduces the
explanation bane ico konna fu ni sum “ move their wings like this ”.
In this example tori is the subject, but. in the sentence Tori wa tobu
toki no shisei wo mitamae. tori wa is not the subject but has the meaning
tori w'o ba (YY., pp. 50-1).
If, as Yamada suggests, Tori tea tobu toki which means “ Birds,
when they fly (or when flying) . . .” is incomplete, so is Tori ga tobu toki
which signifies " When a bird flies . . .” In the first instance the
predicative element is lacking, while the latter is part of a compound
sentence of which the second co-ordinate clause is not filled up. Y hen
therefore Tori wa tobu toki is supplied with a predicative element like
hane wo konna fu ni sum, the resultant Tori wa tobu tola {ni) hane u'o
28
S. YOSHITAKE
I'onna fu ni sum is a complete statement convepng the meaning
'■ Birds, when flpng, move their wings like this or “ When birds
fly, they move their wings like this We can likewise make Tori ga
tobii toM complete by filling up the defective element with, e.g. kuki
ga vgoku “ the air is stirred thus Tori ga tabu toki (ni) kuki ga ugoku
“ "When a bird flies, the air is stirred Further, the sentence Tori
ica tobii toki no shisei ivo mitamae represents Tori (no baai ni) wa (sore
ga) tobu toki no shisei ivo niitamoe ‘‘ In the case of birds, watch their
posture when they fly ! and presupposes a previous statement like
TJo ya tori no kodo ivo kansatsii sum no wa hijo ni kyomi ga fukai.
Mazu no ga donna ni oyogu ka wo mitamae “ It is most interesting to
observe the movements of such creatures as fish and birds. First,
watch how fish swim. ’ Consequently tori wa “ in the case of birds ”
is not the object of mitamae “ watch ! ”, as Yamada believes, but stands
for a suppositional clause ; the direct object of the verb is undoubtedly
shisei “ posture ”, If, however, we change wa in the above example
into ga and say Tori ga tobu toki no shisei wo mitamae “You watch the
posture of birds when they fly ! ” in such a connection as “ Look how
gracefully that aeroplane is going along ! Yes, but you just watch
how birds fly ! They glide just as gracefully ”, tori ga will become a
part of the direct object of mitanuie “ you watch ! ”
In an attempt to press his point that wa does not necessarily indicate
the subject of a sentence Yamada quotes three examples of the con-
struction : Hito no kokoro koso utate am mono wa are “ How strange
a thing is the heart of man ! ” (GM., p. 250 ; WT., p. 292). Explaining
this example, he states that whereas hito no kokoro “ the heart of man ”
is here the subject, wa is used after the predicative utate am mono
“ strange a thing " with the signification of the modern de (YY.,
pp. 55-6). He is certainly right in treating utate am mono as a
predicative (or a complement), but surely wa can never acquire such
a meaning ; we much here supply nite after mono, and it is this nite
that carries the force of de.
Yamada proceeds to dwell on a particular use of koso in which
the verb put in the perfect form (Izenkei) does not refer to the element
to which koso is affixed. Citing four examples like Chichi mikado no
kurai ni tsukasetamaite itsuka to iu hi ni umaretumaeriken koso ika ni
ori sae hanayaka ni medetakariken to oboekabere “ I cannot help thinking
how bright and joyous an occasion it must have been, seeing that he
was born on the fifth day subsequent to the accession of the Emperor,
his father " (OK., p. 115), he points out that it is the verb oboehabere
-V'<
29
i
THE JAPANESE PARTICLES \VA, GA, AND MO
that is expressed in the perfect form, and not medetaJcariken which
refers to the group of words preceding Itoso. He ascribes this con-
struction, which is in his opinion an abnormal structure in spite of the
fact that it was extensively used during the j\Ian-y6 period as well as
in the Heian epoch, to a transilient influence of the particle which has
a great power of governing the predicative element that follows it
(YY., pp. 56-7).
From this simple exposition we learn nothing, for he does not tell
us why the perfect form of a verb is u.sed in conjunction with koso
and what force it has when so employed. When, however, we compare
the above sentence with a somewhat similar instance like Zukyo
nado wo koso tea su nare Certain scriptures should be read " (GM.,
p. 98 ; WT., p. 117), where the verb su is put in the conclusive form
(Shushikei) and nare in the perfect form, we can perceive that the
particle koso in these cases has to do with the final verb and not with
the one that goes before it. It is plain that in this latter example
nare together with koso governs the remaining part of the sentence
with the meaning ” It is proper that . . .". Thus Waley's translation
is a more readable way of putting “ It is proper that we do such things
as reading scriptures ", which is the literal signification of the sentence.
Accordingly, in the example quoted by Yaraada, the final verb together
with the particle koso modifies the rest of the sentence with the meaning
“It is but natural to think that . .
The above review, brief though it is. exhausts the salient points in
Yamada's argument on kakari and tnusubi. If, as he maintains, wa
requires a certain statement, so does ya, which is distinguished by him
as one of the " case auxiliary words’’ from the class of particles he
calls ■■ kakari auxiliary words ". What is really meant by the users
of the words kakari and musubi seems to be that u'u, mo, koso, and
certain other particles imply that the element which follows any one
of them is bound up by the element which immediately precedes it.
Translate, if you like, kakari "binding’’ and musubi "bound up",
but since these terms apply to nearly all particles nothing is gained
by introducing them into the already complicated study of Japanese.
What, then, are the functions of the particles under consideration ?
Yamada states that wa has the signification of exclusiveness and is
used to designate a thing clearly and to prevent its being confused
with other things, while ya serves to modify a noun or noun-equivalent
or to indicate the subject of a sentence (YB.. pp. 253. 207). This view,
which is shared by Kiyeda (KK.. pp. 551. 464), is quite inappropriate
30
S. YOSHITAKE
because in sentences like Kore wa watakushi-no desu “ This is mine ”
and Kore ga ivatakushi-no desu This is mine ’’ ^ both ica and ga
indicate the subject and are alike used to designate a thing clearly and
to prevent its being confused with other things.
Equally unsuitable is Sansom's interpretation. According to
him, u'a serves to relate subject and predicate of a logical
proposition”. ‘’It is separative or emphatic to this extent,” he
continues, “ that the mental process by which any logical proposition
is formed consists of two stages, first an analysis and then a synthesis.
\Mien we say ‘ fire is hot ’ we have first selected from all the concepts
in our minds the particular concept ‘ fire ", and then we predicate of
it some selected property. Wa in Japanese denotes the concept selected.
It might thus be called selective, separative, or distinguishing ”
(SH., p. 258). But when anyone says ” I say, John was here this
morning ”, where John ” would be followed by ga in the Japanese
equivalent, we must consider that the speaker has first selected the
particular concept John ” and then has predicated of the person by
adding was here this morning ”. All that Sansom here says of wa,
therefore, is applicable also to ga, which, following him, indicates the
subject of a sentence (SH., p. 233).
When we turn to Chamberlain we find that he, too, throws no
light on this point. He tells us that " ga is used as a sign of the
nominative case (CH., p. 66), while mi ” is now used as a separative
or isolating particle, corresponding in some measure to the French
' cjuant a ' ”, and serves to lift the preceding element out of the regular
current of the sentence and set it in a place apart, as in the use of the
French word " lui ” in a construction like " Lui. qu'est ce qu'il en
dit ? ” (CH., pp. 85-6). Further, speaking of the difference between
ira and grt, he asserts that when a speaker has in his mind a predicate
and gives it a subject, he uses ga, but when the subject is uppermost
in his mind and he gives it a predicate he uses ira " (CH.. p. 89).
This interpretation, which is .supported by Ro.se-Innes (RC., pp. 82-3)
and Matsu.shita (cf. KK.. pp. 55.3-4), is in a way correct if we were to
accept the definitions usually given of ” subject” and “ predicate ”
in grammars. But when we read Chamberlain's final remark that
tea is emphatic and separative ” (CH., p. 91). in spite of his a.ssertion
that the use oiga necessitates empha.sis on the subject in the English
translation, whereas the use of ira necessitates emphasis on the
' In this article words that are stressed in the English sentences are printed in
italics.
THE JAPANESE PARTICLES WA, GA, AND MO
31
predicate ” (CH., p. 90), we are driven to close the book and sigh in
despair.
Relief is brought by Noss, who explains that “ ga simply marks
out the subject, excluding other things, while wa indicates that an
important predicate is to follow ” (NT., p. 3). Still more lucid
definitions are given of these two particles by McGovern, according
to whom “ ua serves to emphasize the predicate, and ga the subject ”
(MC., p. 15). That neither of these two explanations is perfect will be
seen when we try to apply their definitions of u'a to a construction like
Kore wa watakushi-no de wa arimasen “ This is not mine for their
interpretation does not account for the use of the second wa which is
found within the predicate.
In a sentence like WatakusM wa tahakoga dai-suki desu I am very
fond of smoking ” Chamberlain believes that ga preserves its older
meaning “ of (CH., p. 65). It is true that the words suki " fondness ”,
hoshii “ be desirous ”, iru “ be in want ", etc., usually require ga
before them, when the particle corresponds very closely to "of”.
However, such instances as Wa ga tni ga hoshi kuni va " The country
I fain would see " (K., p. 210 ; CK., p. 331), where mi ga hoshi means
“ desirous of seeing ", are rare in eighth-century literature. Instead
we usually find the words, which would now bo followed liy ga in
similar constructions, standing alone without any particle. For
example : Kamitsuse wa se hayashi " The water in the upper reaches
is (too) rapid ” (K., p. 26 ; CK., p. 48) would be Joryu wa nagare ga
hayai in modern colloquial, and Sore kao yoshi ‘‘ She is good-looking ”
(K., p. 185) answers to modern Sono onna wa kiryo ga ii. Even in the
language of the thirteenth century ga was not used in such a position,
e.g. Onna %va chikara yowashi " Women have little strength ” (U.,
p. 270), in expressing which the modern Japanese would use ga and
say Onna wa chikara ga yowai. On historical grounds, therefore.
Chamberlain's explanation is imperfect, even if it is not inaccurate.
Moreover, it is only applicable to certain cases.
Quoting an example Kitsune wa o ga nagai “ The fox has a long
tail " Noss declares that the subject with wa is here " grammatically
disconnected from the sentence, while the predicate nagai (long) takes
the subordinate subject o (tail) with ga (NT., p. 6). He goes on to
say that in a sentence like iSihon wa yama ga 6i " Japan is
mountainous" (lit. "In regards to Japan, mountains are many”,
according to him) " grammatically yama ga 6i is a complete sentence,
but the expression simply fills the place of an adjective (NT., p. 6).
32
S. YOSHITAKE
On this particular use of ga Yoshizawa has written an article
(YT., pp. 1-10), in which he states that in expressions like Ji ga kakenu
" I cannot wTrite ”, Hon ga iru “ I am in want of a book ”, Kashi ga
suki da " I am fond of sweets ”, Mizu ga nomitai “ I want to drink
water ", Kane ga hoshii “ I want money ”, and Hebi ga kowai “ I am
frightened of serpents ”, ga is generally regarded as indicating the
grammatical object. Refuting this prevalent opinion he contends that
the idea Hebi ga kowai cannot be expressed in the form Hebi wo kowai
by using wo (which is a sign of the accusative case) any more than is
it possible to represent Hebi wo kowaku ornou ” I am frightened of
serpents ” by Hebi ga kowaku omou. Further, we cannot simply remove
the -tai element from Mizu ga nomitai and say Mizu ga nomu, whereas
if we retain the same element we cannot say Mizu wo nomitai by
replacing ga by wo. That this form is occasionally found in novels
is due to the writers' wrong analogy on Mizu wo nomu, etc. It follows
from this that whilst wo connects mizu “ water ” and nomu “ drmk ’’
ga binds mizu and -tai ” wish ”, aided by another element nomi
“ drinking ”. For these reasons Yoshizawa, in agreement with
Mitsuya, considers the words designated by ga in the above examples
as the subjects of the respective sentences.
It must not be hastily concluded, however, that by using the word
“ subject ” Yoshizawa means that the word to which ga is affixed
in the cases under consideration is the subject of the whole sentence,
for he intimates that in the above examples the ” principal (or general)
subject ” (Soshu) such as e.g. watakushi icu I ” is omitted.
In order to explain the difference between these two kinds of subjects
he cites two more examples ; Tokyo wa jinko ga di Tokyo is
populous ■’ and Chojiwhugyo mina sei ari “ Birds, beasts, insects
and fish are all sexual ”. ” In these sentences," he says, " everyone
will treat jinko ‘ population ’ and sei ' sex ’ as the subjects of the
respective sentences, as it is only right to do so."’ “ Just as,” he
continues, “ Hebi ga kowai is used in the sense of Hebi wo kowaku
omou, so the sentence Tokyo wa jinko ga di has the meaning Tokyo
wa jinko wo oku motte iru ' Tokyo has a large population In the
same way, sei ari in the above example means, in his opinion, sei ivo
motte 'iru “ have se.x , not sei to iu mono ga sonzai shite iru ‘‘ what is
called ‘sex’ exists '. Thus he concludes that in these sentences
Tokyo and chojuchugyo are the “principal (or general)” subjects of
the respective sentences, and jinko and sei the “ auxiliary ’ subjects.
This dissertation by Toshizawa is unfortunately devoid of value,
THE JAPANESE PARTICLES WA, GA, AND MO
33
for at least two distinct constructions are mixed up in his treatment.
If, in the sentence Ji ga l:ahenu ‘‘ I cannot write ”, the “ principal ”
subject such as watukushi tea “I” is omitted, as Yoshizawa
suggests, then in the expressions Kato tea Eigo tio lion ga yonieru
“ Kato can read English books ” and Kato ni tea konna Jion tea
yomenai “ Kato can't read a book like this ’ the word Katd must
likewise be the main subject. But in the last example Kato is followed
by ni which can never be used with any subject, "principal' or
‘‘subordinate”, in the grammatical sense of the word “ subject
This means that Kato in these sentences is not the proper subject
and that the particle ni, which is affixed to Kato in the negative
expression, is left out in the former construction. On the other hand,
we cannot say Boku ni tea mizu ga nomitaku nai for Baku tea mizii ga
nomitaku nai. Accordingly, the -tai construction, together with such
cases as Kane ga hoskii, Kashi ga siiki da, and Hebi ga koieai hav’e to
be explained differently. Then again, the idea Chojuchfigyo mina sei
ari may be expressed in the form Chojiichiigyo ni wa mina sei ga am.
where ni is used after the word chojnchugyo which Yoshizawa regards
as the main subject. Since in this last sentence the particle ni has
a locative function, the word ckojuchdgyo cannot be a subject of any
kind. As a matter of fact, Yoshizawa has not missed this point, for
he adds that in the sentence Koppu ga bon ni nosete arimushita “A
tumbler was placed on the tray ”, which may stand for Bon tea koppu
ga nosete arimashita “ The tray had a tumbler (placed) on it ”, the
word bon “tray” is expressed in a different capacity and cannot
therefore be repeated as representing the “ principal ” subject of the
sentence. It is obvious that Yoshizawa was then at the end of his
resources. He would have been saved from offering this unsatisfying
explanation if he had realized that in the last sentence the particle
ni, which might have been affixed to bon, had been dropped. But
since this use of ni differs from that of the same particle in Kato ni tea
konna hon tea yomenai. the construction Chojuchugyo mina sei ari
requires separate consideration.
The problem has also been dealt with by Yamada. According to
him. when a sentence contains two subjects, one of them indicating
a part of that which is denoted by the other, the subject which
designates the whole is the “ principal ” subject, and the subject which
represents a part is the “ subordinate ” subject (YB., p. 454). Thus, in
* In this article I have been compelled to use several terms like “ subject ”, etc.,
which are both ambiguous and unsatisfactory.
VOL. Vltl. PAET 1.
3
34
S. YOSHITAKE
the sentence Zo wa Tcarada ga oMi “ The elephant has a large body ”
the word zo “ elephant " is in his opinion the principal subject and
karada “body” the subordinate subject. But in another example
given by him, namely Shzu wa iro ga gin ni nite iru “ Tin resembles
silver in colour”, can we consider iro “ colour" as a part of suzu
“ tin ” ? This niay be accepted on the ground that colour is one of
the attributes of the metal, but it would certainly be impossible to
say that, in the example quoted by Chamberlain, tabako “ tobacco
(or smoking) ” is a part of watakiishi “ I ”, although such a figurative
locution as “ Music is a part of me ” is quite common in English.
To make the matter more complicated Yamada distinguishes his
“subordinate .subject" from what he calls chinjutsuku “predicative
phrase ". His contention is that in the above examples Zo wa karada
ga okii and Suzu iva iro ga gin ni nite iru, the words karada ga and iro
ga merely serve to add explicitness, and therefore the sentences would
be intelligible without them, but that in expressions like Tokyo wa
mkura ga di “ There are many cherry-trees in Tokyo " and Ano liito
wa komi ga umai “ He is sociable ", if the words sakura ga “ cherry-
trees " and kosaiga “ social intercourse " were omitted, the meaning of
the sentences would be either altered or lost entirely (YK., pp. 387-9).
However, before acknowledging such an interpretation we must
pause to think how we actually speak. Suppo.se a Japanese says Boku
iku. Is it a complete statement ? As an answer to the question Kimi
ikii n' kai " Are you going ? ” it has a complete meaning '• I am ”.
But it is undoubtedly incomplete as a chance statement, for it does
not tell whither, how, when, or why the speaker is going. And this in
spite of the fact that it consists of a .subject boku " I " and a predicate
ikn " go " that requires no direct object. In point of fact, our utterance
is almost always incomplete, and this incompleteness is the beauty of
language, for it appeals to the listener's intelligence and arouses his
imagination. How unbcarablv monotonous and irksome laniniace
would become if we exprcs.sed more than we are accustomed to !
Who would be patient enough to listen to a conversation of this kind :
■■ Have you been to .see the show you said you were going to the la.st
time we met ? Yes. I have been to see the show I said I was goin" to
the last time wc met " 1 To us who converse bv word of mouth the
Japanese saying" Bett('r defect than excess" is indeed a merciful advice.
Thus Tokyo wa di (without mkura ga) and Ano Into wa umai
(without kdsai ga) perfectly express the meanings “ There are manv
cherry-trees in Tokyo " and " He is sociable " when thev are uttered
THE JAPANESE PARTICLES WA, GA, AND MO
35
subsequently to,e.g., Kobe tva saTcura ga sukunai “There are not many
cherry-trees in Kobe and Bokii ica Icdsai ga beta da “ I'm not
sociable If Tokyo wa oi when uttered of a sudden does not convey
any definite meaning, so would Zo wa okii be ambiguous to a person
who has never seen an elephant. It is because we know something
of the elephant that this expression immediately suggests to us that
the speaker is referring to the elephant's body and not to its eyes or
its ears. If, therefore, anyone speaks abruptly of a Hr. Kato whom
we have not met and says Kato-mn wa okii we shall fail to understand
whether he is speaking of Kato's physique, his head, his eyes, or some-
thing else, because this expression may stand not only for Kato-san
ivu karada ga okii “ Mr. Kato is big-hodied " but also for Kalo-san
wa atama ga okii “ Mr. Kato has a large head ", Kato-san wa me ga
okii “ Mr. Kato has large eyes ", and so forth, according to the contexts
of situation.
There is, in fact, no need to make such a fuss about so simple a
matter as the one in question. M’hen we compare Kore iva hinshitsu
ga warui “ This is inferior in quality" or Kore wa gaiken ga warm
“ This is bad in appearance ’’ with Kore wa ichibuhun warui “ This is
partly bad ” or Kore ica dono ten iii oite mo warui “ This is bad in every
respect ", we can at once recognize that the element which immediately
precedes warui " is bad " in each of these sentences modifies the
meaning of the adjective. It is therefore an adverb equivalent. Thus
the particle ga affixed to kashi “ sweets ", mizu " water ", kane
“ money ”, hehi “ serpents ", and jinko " population in the examples
quoted by Yoshizawa, and to the words karada " body ', ?ro “ colour ”,
sakura “cherry-trees", and kosai "social intercourse" in the sentences
cited by Yamada, serves to restrict the application of the word or
group of words which follows it to the one that immediately precedes
it. Its function is in no way different from that w’hich is found in
certain set-phrases in modern Japanese, e.g. wa-ga-mama " way-
M-ardness ”, sore ga tame {ni) "on account (or in consequence) of that ".
It is with this force that came to be used with the so-called “subject''
of a sentence, as in Kore ga watakushi-no desu “ This is mine , Kyd
ga toka desu " To-day is the tenth ”, Kore ga ii “ This is good (better,
or the be.st) ”, and Watakushi ga ikimasu " I am ’ (in answer to the
question “ Is anyone going ? ").
Historically speaking, the use of ga in the context under con-
sideration seems to have been confined to a noun-phrase or noun-
clause in the eighth century, as may be seen from the Kojiki passage
36
S. YOSHITAKE —
Wa ga mi ga TiosJii hmi ica cited on p. 31 above. Precisely when this
use was extended to other cases I am not in a position to assert beyond
stating that the modern uses can be traced back to the fifteenth century
(for examples cf. YM.. p. 259). It is well to note that in about
A.D. 1400 the particle ivo was used in the -tai construction, e.g.
Kuwashiki mime ivo kikitaku wa ‘‘ If you want to know the details "
(G., p. 449), which corresponds to the modern Kuwashii koto ga kikita-
kereba. Some Japanese still prefer this use of wo, though disapproved
bv others like Yoshizawa. Even those who use ga in Mizu ga nomitai
would not hesitate to employ wo in Mizu wo nomitai to omou “ I feel
like having a drink of water". Besides, the form -tagaru takes wo, not
ga. The point is this : no matter whether it is accompanied by ga
or wo, the word mizu ” water when used in conjunction with nomitai
“ want to drink " and preceded by hoku wa. is a part of the predicate.
With regard to the “ emphatic particle " tea in contrast with 20
and koso Sansom states as follows : ” Emphatic particles are freely
used in Japanese, for two very good reasons. In the first place spoken
Japanese has an even accentuation, and it is therefore not easy to
emphasize words by oral stresses. In the second place English, for
instance, has other ways of showing emphasis, which are not available
in Japanese. Thus we can say ’ I did go ’ instead of ‘ I went ’, or we
can say ‘ John it was ' instead of ’ It was John ' ; but Japanese
does not allow of such modifications or changes in significant word-
order. These considerations go a long way towards explaining the use
of emphatic particles where emphasis is required, but they do not
sufficiently account for all uses of wa. For one thing, if wa is emphatic,
so are 20 and ko.so. and there is not much difference, except in degree
of emphasis, between hi wa atsushi. hi zo atsnki, and hi koso atsukere.
Seeing that all these particles existed in a relatively primitive stage of
the language, it is surely unlikclv that the language would have
developed such a refinement as these grades of emphasis unless forced
to it by a deficiency in some other direction " (SH., pp. 257-8).
How inadequate Sansom’s observations are can easily be seen when
we compare the following four expressions ^ : —
(1) Anala {wa) ikimashita ka? = Did you go ?
(2) Anata {wa) itta n' desii ka? = “ Dul you go ?
(3) Anata wa ikimashita ka ? = Did you go ? ”
(4) Anata ga itta n desu ka ? = Is it you who went ? "
^ In this article words that are stressed in the Japanese sentencf^ are printed in
Roman letters.
THE JAPANESE PARTICLES WA, GA, AND JIO
37
The plain fact is that Japanese does make use of oral stress, and
Wataliushi (u'a) itta ir dem “ I did go ” is distinguished from
Watakushi (iva) ikimashita “ I went Further, “ John it was who said
that ’’ ^ would be Jon da so itta no u'a as a contrast to So itta no im
Jon da “ It was John who said that Inversions of this kind have
always existed since the earliest period of the Japanese language of
which we have knowledge.
The functions of zo and koso are entirely different from that of ira.
In ancient Japanese the subject of a sentence was not marked bv anv
particle when it was thought that the question of emphasis could be
gathered from the context. But when it was felt necessarv to show
that the subject was emphasized, zo was affixed to it, whilst the especial
emphasis on a predicative word was indicated by using ira before it.
Thus Hi at sushi may have meant '' Fire is hot " or ” Fire is hot "
according to circumstances, while Hi ira utsushi " Fire is hot " and
Hi zo atsuki “ Fire is hot (hotter, or the hottest) ” had more dehnite
meanings. Towards the end of the tenth century the nominal con-
struction Hi ga atsuki (koto) "Fire's hotness " was about to be
supplanted by the Hi ga atsushi " Fire is hot ". and since that time the
latter form gradually took the place of the other nominal construction
Hi zo atsuki which appears to have once meant " fire itself hotness
The following examples will suffice to show the sequence of
development.
(1) /rose zo hashiki " Mine elder brother is dearer ” (K.. p. 138 ;
CK., p. 226).
(2) Wa ga kataehi ivo kakimamitamaishi ga ito hazukashili
koto “ Thy having peeped at my (real) shape (makes me) very
shame-faced " (K., p. 94 ; CK.. pp. 152-3).
(3) Kaku icarai imasuiu ga hazukashi " Your ridicule . . . makes
me feel embarrassed " (MS., p. 668).'^
(4) Mizu no soko ye kashim u'o irete wireba hontai ga nai " AVhen
he ducked his head into the water he found that what he wanted
was not there " (SA., p. 91).
The particle koso seems to have a very interesting history behind it.
One of its earliest u.ses was to designate wish, when the particle followed
* Such a form as " .John it was ’’ never stands alone in Engli.sh ; it is invariably
followed by a noun-clause.
- In each of the examples (2) and (3) it is hard to decide whether the element
preceding ga is the subject of the sentence or a part of the predicate. .All that we can
safely state is that the group of words to which ga is affixed restric ts the application
of the adjective that follows it.
38
P. YOSHITAKE —
the adverbial form (Renyokei) of a verb and stood at the end of a clause.
This particular use appears to have already been obsolete, or at least
obsolescent, in the spoken Japanese of the eighth century, being
preserved only in the language of poetry in that period (for examples
see OM., pp. 243, 263-4). According to the current reading, the
language of the Kojiki contained various uses of koso, the commonest
of which may be illustrated by the following examples : —
(1) Are tsune iva umitsitji u'o toshite kayoican to koso omoishi
ICO " I had wished always to come and go across the sea-path.
But . . . ■’ (K., p. 94 ; CK., p. 152).
(2) A ga moit tsuma ari to iiiaba koso ni ie ni mo yukame kuni
ICO mo shinubame “ If they said that the spouse whom I love were
(there), I would go home, I would long for my country ’’ (poem,
K., p. 237 ; CK., p. 366).
(3) Sono toki ni koso are kanarazu ai-iwame ‘‘ At that time I will
surely meet and speak with thee ” (K., p. 226 ; CK., p. 350).
(4) Ubeski koso toitamuc ” It is indeed natural that thou
shouldest deign to ask ” (poem, K., p. 220 ; CK.. p. 344).
(5) Na koso ica yo no nagabito Thou indeed art a long-lived
person ” (poem, K., p. 220 ; CK., p. 343).
(6) Rare kono tachi wa tatenintsuru ni koso “ So I just present
this cross-sword to thee " (K., p. 101 ; CK.. p. 164).
(7) Kono shiroki i ni nareru mono mi sono kami no tsukaimono
ni koso arame “ This creature that is transformed into a white
boar must be a messenger from the Deity ” (K., p. 161 ; CK.,
p. 262).
(8) Are mono iicazu tada iita iro koso utaitsure “ I said nothing ;
I was only singing a song " (K., p. 132 ; CK., p. 217).
Yamada's simple e.xplanation that koso points out something
pre-eminently ” (YB.. p. 258) is no e.xplanation, for his definition
of ica cited on p. 29 above amounts to the same thing. Chamberlain
treats koso as an emphatic particle pure and simple, rendering its
meaning by ■'surely", “ indeed etc. But when we observe its
frecjuent association with verbs ending in -me. the perfect form (Izenkei)
of the probable mood, we cannot help suspecting that there must be
something deeper after all in the function of koso. Although
Chamberlain’s translation of example (1) is almost correct, we may.
for the purpose of bringing out the force of the particle more distinctly,
translate it more literally thus I had always wished that I would
(be able to) come and go by the sea -path, and now. . We shall then
THE JAPANESE PARTICLES WA, GA, AND MO
39
be able to see very clearly that koso here serves to indicate an unrealized
wish, for the expression of which the subjunctive mood is used in
English. In example (2) the same particle introduces two subjunctive
clauses. Example (3) really means “ If you do so, then I would
certainly talk with you ”, and therefore koso here again introduces
a subjunctive clause. So does it also in (4), as plainly shown bv
Chamberlain's rendering. The same scholar’s translation of (5) gives
us the impression that the particle served merely to emphasize the
statement. But when we compare this example with a modern
expression like Rare koso shin no aikokusha da ” If ever there has been
a true patriot, he is one ” we are led to conclude that the Kojiki
passage must have the meaning “ If ever there has been a long-lived
person, you are one ". It is true that both English and French, for
in.stance, use the indicative in “ if ever ” clauses, but since the position
of the word “ ever ” here intensifies the doubt, we may consider that
koso introduces the subjunctive, ‘‘ you would be one,’" also in the
present construction. The force of koso in (6) is intensifying in all
appearances, but if we translate the sentence into “ So I will (or wish
to) present this cross-sword to you ", which is equally suitable for the
context as Chamberlain's interpretation, we shall find that here once
more the particle is used in a subjunctive connection, because the
subjunctive is the mood of ” will ". The translation ” must be ” for
koso arame in example (7) clearly shows that the particle here serves
to indicate the speaker's judgment upon the state of a third person.
The use of koso in (8) can onlv be regarded as emphasizing the
statement. So we can see that already at the beginning of the eighth
century the particle koso had a variety of uses. But how did all this
come about ?
In the history of the majority of the European languages the
optative has been very closely related to the subjunctive, the former
being the mood of " wish " and the latter the mood of " will ", so much
so that nearly all subjunctives in the Germanic languages may be
traced back to optatives (cf. S.S., pp. 63-6, 85-109). The subjunctive
itself has been encroached upon by the indicative in some languages,
as, e.g., in French " si j'etais la while both English and German still
keep the subjunctive '' were " and ” ware ” in like circumstances.
Such being the case with many languages, it would not be
unreasonable to conjecture the existence of similar phenomena also
in Japanese. It is not unlikely that from the verb kosu “ wish ’ there
evolved the particle koso, which first served to denote wish (cf.
40
S. YOSHITAKE
TK., p. 157 ; YA., p. 666).^ Thence its use was extended to the
contexts which require subjunctives in some other languages. If the
subjunctive is the mood of “ will the principal duty of koso used in
such a context must be to express volition, i.e. the speaker's
determination to doing or not doing something. From this the use of
koso could be further extended to an expression of determination
concerning the action or state of someone else, i.e. judgment or con-
clusion, and the transition from decisiveness implied by conclusion to
emphasis is not difficult to imagine. This assumption seems to explain
all the principal uses of koso in the Japanese language of the eighth
century. Incidentally it also accounts for the close relationship between
koso and namo (> nan), which latter was used for designating volition
(hence intention) and desire (cf. MG., pp. 258-262).
Such a series of developments in the function of koso could not
possibly have been realized in a short space of time. In other words
the particle in question, hence also the perfect form (Izenkei) of verbs
which was associated with it, must have had a comparatively long
history already 1,200 years ago. In face of this likelihood Sansom
asserts that the perfect form came into use in Japanese just before
the Kara period " on the dubious ground that if it " had existed in the
language from which both archaic Japanese and Luchuan are descended
it would have left .some traces in Luchuan " (SH., p. 143). The fallacy
of this inference is shown by Ifa's observation that Luchuan had at
one time a word or particle, pronounced su {sha, sko, ot jo), which
corresponded to Japanese koso and which required the perfect form of
verbs as did koso (IK., pp. 406-413). In my opinion the .so-called
“ honorific ” suffix -su in ancient Japanese originallv served to denote
volition (cf. ALA.., p. 654), and it is probably to this Japanese suffix
that the old Luchuan sit is related. If this hj'pothesis be granted, then
we can state that in Luchuan sn survived in the forms sit, shii, sko,
and jo until about the seventeenth century, but that in Japanese it
had already been absorbed by the ninth century in the old optative
koso, which has been handed down to us with a comple.xity of uses.
A systematic study of the exact relationship between the various
duties of koso and the positions which the particle occupies in sentences
under different circumstances would enable us to establish the true
sequence of development, but a detailed discussion of such a matter
lies outside the scope of the present article. AVe must content ourselves
^ In this use Icoso may have retained its verbal nature, as it was regularly preceded
by thf2 adverbial form of a verb.
THE JAPANESE PARTICLES WA, GA, AND 110
41
■with but one particular instance, in which koso is employed for an
expression of request with the meaning " I pray ” and is afiixed to
the name of a person. This use of the particle seems to have come
into being some time during the tenth century, undoubtedly as an
extension of its older function of designating wish. For example :
Ve koso. Kono tera ni ouaseshi Getiji-no-Kimi koso oicasMian nare.
Nado mitamau'anu “ Pray, Grandmother ! Prince Genji who came to
see us in the mountains is here, pajdng a visit. Why do you not let
him come and talk to you 1 (GM., p. 140). Waley translates Ve koso
“ Grandmother, Grandmother ! ” (WT., p. 163) and Ukon-no-Kimi
koso “ Ukon Ukon " (WT., p. 102), but it is manifest that koso here
means “ I pray ”. Matsuoka, too, is inaccurate in stating that in the
above example a word like mitamae “ look ! ” is omitted after
Ve koso (MG., pp. 224-5).
Not a single ■writer on Japanese grammar has attempted to explain
the use of wa with the direct object of a verb. The direct object is
indicated by wo only under two conditions : (1) When the speaker
presumes that the object is not in the hearer's mind ; and (2) when the
speaker wishes to draw the especial attention of the hearer to it. When,
however, the speaker thinks that the object is more or less apparent
to the hstener he would use wa in place of ivo. To take an e.xample,
in answer to the question Anata («r«) kono hon wo mimushita ka " Have
you seen this book ? ’’ the Japanese would always replace wo by wa
and say Hai, sono hon iva (nio) ndmaAnta " Yes, I've seen that book "
(or le, sono hon ua mada nriniasin " No. I haven't seen that book ")
if the word hon book ’' is to be repeated. It is quite evident that the
informative element of the answer is mimashita " I've seen " (or nmda
mimasen I haven’t seen "), hence the less informative element sono
hon ‘‘ that book '' is followed by wa when it is expressed. If a Japanese
asks you Kono hon wa mimashita ka “ Have you seen this book 1 "
by using ica in place of wo it is not because he thinks that kono hon
“ this book " should be isolated ” as he has omitted anata wa
“ you ’', but because he is showing you the book, and so he presumes
that it should be obvious to you that he is speakmg of it.
Here it may be noted that iSansom did try to explain this particular
use of tea, but quite unintentionally. He states : “ Probably one of the
best illustrations of the true function of wa is provided by the Japanese
idiom which is commonly used when in English we should emplov
a passive construction. In English a sentence like ' This house was
built by niy father ' is of a normal tj-pe, but the Japanese idiom does
42
S. YOSHITAKE
not favour a passive construction applied to the name of an inanimate
thing, because an inanimate thing like a house cannot get an action
performed, cannot, for instance, get itself built. Consequently in
Japanese the correct rendering of the above sentence is Kono uchi tea
chichi ga tatemashita, viere the subject of the logical proposition hono
uchi ‘ this house ' is designated hytea, and the predicate is the complete
sentence chichi ga tatemashita ‘ my father built ’ ” (SH., pp. 258-9).
That this interpretation is wide of the mark can easily be seen when
we change the order of the words and say Chichi ga kono uchi wo
tatemashita, which has much the same meaning as the sentence quoted
by Sansom. It is clear as crystal that kono uchi in both these con-
structions is the direct object of tatemashita “ built ”. The sentence with
wa really means "You see this house here ? My father built it ” or
“ This house you are in now — well, my father built it ", while the
construction with v:o is generally used in an adverb-clause, e.g. Chichi
ga kono uchi tvo tateta toshi ni watakushi wa Eikohu ye ikimashita
“ I went to England in the same year as my father built this house ”.
A structure like This house was built by my father has no exact
counterpart in normal Japanese, but since this form of expression is
used in such circumstances as pointed out above, the Japanese convey
the same concept in the form quoted by Sansom. If we were to follow
his instructions we should find ourselves in a hopeless situation directly
we venture to translate ‘‘A clock was stolen by a thief”, because
Tokei wa dorobo ga nmumimashita has a different meaning : “ The
clock (mv clock, etc.) was stolen by a thief ". WTien, however, we
reflect upon the possible reason why the English sentence has to be
rendered bv Dorobo ga tokei iro nusumimashita, where tokei “ clock ”
is accompanied not by tea but by tro, we are bound to realize that it is
not the sheer aversion to the application of a passive construction to
the name of an inanimate thing that calls for the use of wa with the
direct object of a verb. It must be the intrinsic function of wa as
designating the relatively better known element of a sentence that has
prompted the use of wa in such a position. Naturally words like “ a
clock”, "clocks”, that have indefinite meanings can never be
accompanied by wa. no matter whether they are used as the subject
of a sentence or as the direct object, unless they have been mentioned
in a pre\uous statement or are used in the generalized sense " the
machine called ' a clock '
There remains to be considered the particle mo. According to
Sansom, “mo may best be regarded as complementary to wa, for where
THE JAPANESE PARTICLES WA, GA, AND MO
43
u'a excludes one thing from other things, mo includes one thing with
other things ” (SH., p. 263), while Yamada defines mo as “ contrasting
and inclusive, indicating a certain thing and hinting at the existence
of similar things ” (YB., p. 255). Sansom’s explanation is not faultless.
Examine the following three sentences : —
(1) Kore ga icatahushi-no desu “ This is mine " (Information ;
Icore “ this ”), in answer to “ Which is yours ? ”
(2) Kore wa ivatahushi-no desu “ This is mme " (Information :
u'atakushi-no “ mine ’'), in answer to ‘‘ Whose is this ? ”
(3) Sore mo ivatakushi-tw desu “ That's mine, too ” (Information :
sore mo “ that . . . too ’), in answer to “ What else is yours ? ”
Anyone upon comparing these three sentences might argue that
mo, when used with the subject of a sentence, is complementary as
much to ga as it is to wa. The definition offered by Yamada elucidates
the reason why mo is affixed to the subject of a sentence whose
predicate has not been mentioned previously when the speaker implies
something else for which the same predicate holds good. Thus, in
a sentence like Nihon mo kore kara wa dandan atsuku narimasu yo
“From now on Japan will gradually get hotter and hotter, you know '''
the speaker has in mind other countries of similar climate, whereas
if he replaces mo by iva the statement would concern Japan alone,
irrespective of other countries.
This is very clear and perfectly explains the use of mo. but I am
not quite sure whether all Japanese entertain the same thought when
they utter the sentence in question. Speaking personally, when I
express such an idea in Japanese, other countries do not enter into
my mind, no matter whether I use wa or mo ; my intention is to refer
to Japan alone without hinting at other countries. None the less.
I do seem to make a distinction between the use of these two particles
also in this particular case. When I use mo I am speaking of Japan
exclusively but in a general and somewhat non-committal way,
whereas the use of wa would make me feel that I am confining my
statement too strictly to Japan alone. It is probably my knowledge
of other countries that induces me to employ nto in preference to wa
in order to evade such a likely answer as ell. Japan is not the only
countrv ”. Be that as it may. the force of mo in the sentence under
consideration seems much weaker than that in e.xample (3) quoted
above.
Both wa and tno may be found immediately after the -te (-de) form
of inflective wmrds. The -te wa (-de wa) form has either a suppositional
44
S. YOSHITAKE —
meaning “if” or a conditional signification “when, as”, while the
form -te mo {-de mo) conveys a concessive idea “ even if, although ”.
Yamada appears to believe that the forms -te wa and -te 7no contain
the conjunctive form (Eenyokei) -te oi-tsu (YB., pp. 162, 239). Sansom,
who shares Yaniada's opinion, calls the -te form sometimes a participle
(SH., pp. 175, 177) and sometimes a gerund (SH., pp. 243, 347),
whereas Chamberlain (CH., p. 165), McGovern (MC.. p. 31), and Rose-
Innes (RC., pp. 14-16) all give it the name “ gerund ”. But since the
form in question, while doing some of the duties of the English gerimd
and participle, does not agree with either of these, it would he best
to call it simply the -te {-de) form. If a participle in English can convey
the meanings ” if’ and “ as ”, e.g. in “ Strictly speaking (= If we
speak strictly) ” and “ Having (= As I have) so much to do at home.
I scarcely ever go out ”, there should be nothing strange in the use of
the -te form in a suppositional or conditional context. AVhat may seem
curious, however, is the use of ica after the -te form in such a connection.
Neither Yamada nor Sansom speaks a word of this particular use of
u'u, hut they both consider that the form -te ba, as in Kaku hkoshi-
meshite ba " If they thus hear (the ritual words) ” (ON., p. 414 ; FR..
p. 62), contains the imperfect form (Mizenkei) -te of -tsu (YB., p. 239 ;
SH., p. 174). Explaining this very example, Motoori states that the
particle should here be read ba, not ira, and that the form -te ba had
the meaning -te araba, differing from that of -te tea with which it later
became confused (ON., p. 414).
There are at least two reasons that support the accuracy of ilotoori's
opinion. In a sentence like Il«/c v:o ba ika ni scyo tote sutete wa
hoboritaawin zo ' What do you want me to do by going up (to heaven)
leaving me behind (as you are apparently tliinking of doing)? ”
(T., p. 32) we can omit wa u.scd after satete or replace mitete wa by
kakn v:a autete "leaving me behind in such a manner” without
seriously altering the meaning of the sentence. But in a passage like
Akckure minaretani Kagaya-Hitae wo yarite wa ikaya omoubeki ” How
sad he would be if I were to let them take away Kaguya-Hime whom
he has been accu-stomed to sec always around him ! ” (T., pp. 29-30)
no such elision or modification is possible. Further, in the same Taketori
Monogatari, from which the above c.xamples have been taken, we find
the following passage : Kono tama tori-ede wa ie ni kaeri ku aa to
notamawasekeri. Onoono ose uketaaiawarite makari ideim. Tatsn no
htbi no tama tori-ezu ba kaeri ku na to notamaeba ” ‘ Don't come home
if vou can't get this jewel ‘ the Lord said. Each of the servants.
THE JAPAXESE PARTICLES WA, GA, AND MO
45
receiving orders, went off. As the Lord had told them ‘ Don't come
home if you can't get the jewel on the neck of a dragon ' " (T., p. 16).
It will be noticed that tori-ede ( < tori-ezute) wa is here used with
precisely the same signification as tori-ezu ba " if you can't get ’'.
If, as Yamada and others maintain, the form ezu ba contains the
imperfect form -zu (YB., p. 2.39),^ why .should -te in ezute ira, from which
ede ica was evolved, be the conjunctive form and not the imperfect
form ? If the form -te wa used in the present context goes back to
-te ba, and if this ba was invariably used not with the conjunctive
forms but with the imperfect forms of verbs and adjectives (cf. YA..
p. 662), as well as with the imperfect forms of the suffixes -yni (thus
-na ba, cf. SH., p. 182), -hi (thus -he ba, cf. SH.. pp. 183-4). and -zu
(thus -zu ha, cf. OM., pp. 95-6), there is no reason why ica, if this is
a variant of ba, should be construed with the conjunctive form only
in the case of -tsu.
The problem is : Did this -te wa in ancient Japanese really contain
the imperfect form -te of -tsu, and. if it did. when and how it became
confused with -te wa that was built on the conjunctive form -te ? "We
know the imperfect form -ke of the adjective suffix -ki was already
obsolescent in the eighth century, being supplanted by -ku that was
homoplionous, or possibly even identical, with the conjunctive form
-ku. To this change must have contributed, among other causes, the
above-mentioned remarkable tendency which a participle may develop,
because the conjunctive form of adjectives includes the idea ‘‘ being
It is no wonder then that in the ninth century the form -te ba should
have begun to be confounded with -te wa that contained the conjunctive
form -te, wffiich was, and still is, a sort of a participle. But since the
suffix -ten, which served to denote willingne.ss. intention, inclination,
etc., and which was the only other structure built on the imperfect
form -te, was in current use in the spoken language for some centuries
to follow, had there been a linguist, say. in the eleventh century, the
form -te ica derived from -te ba would have been distinguishable to
him from -te wa that contained the conjunctive form -te, just as the
gerund and the present participle are identical in form in modern
English yet can be differentiated in their functions. It is even possible
that the form -te wa which included the imperfect form -te was actually
pronounced -te ha in the earlier part of the Heian epoch but was written
^ Sansom (of. .SH., p. 191) does not seem to recognize the imperfect form -zu.
He must therefore be of the opinion that yuknzu ba if he does not go ’ {SH., p. 194)
contains the conjunctive form -zu.
46
S. YOSHITAKE
with two Kana signs which now stand for te and ha, there being no
diacritical marks used at the time for distinguishing the syllable with
a voiced consonant from the one that had a voiceless consonants
The form -ten , as far as I can trace, sur\uved until towards the end of
the fourteenth century, as e.g. Tadaima sayo no ryori tsuJcamatsunten
ya “Would you now prepare a fish-dish like that! ” (MK.. 623).
But with the disappearance of this suffix in the fifteenth centurv the
confusion between the two -te wa forms would naturally have been
completed. Thus the form -te tea occurring in the passage Saredo mo
kokoro-yowaJcute wa kanaubeki ni arazareba ^ “ But stiU, as it was no
use being faint-hearted ..." taken from the Gikeiki would probably
be one of the forerunners of the modern use of the conjunctive form
-te in comhination with u-a in a suppositional or conditional context.
With regard to the form -te mo Yamada writes as follows : “ Mo
is chiefly used in the spoken language and, joined to the Renyokei
(i.e. the conjunctive or adverbial form) -te of -tsu or the case auxiliary
word de, indicates a supposition that leads to a contrary result. This
use of mo appears to have developed from that of mo in to mo (do mo)
which is used in the written language. Originally mo had no signification
of opposition, but as its use in combination with to (do) continued over
a long period of time, the particle began to be considered as having such
a meaning, and thus came to serve for designating contrariness
(Y’B., p. 245). It is true that no instances of the -knte form of adjectives
followed by mo seem to be found in early Japanese texts. But a passage
like le wo idete mo matsungofo wo okonau ni ani sawarubeki mono ni
wa arazu “ If one were to enter priesthood, that should not interfere
with one's conducting the affairs of the state ” (>S.. p. 162). taken from
the Imperial edict of a.d. 764, would he sufficient to deny his
h^.’pothesis. As in the case of -te ira. it is quite likely that the form
-te mo in the context under consideration contained the imperfect
form -te until towards the end of the fourteenth centurv. In later times
it must certainly have shared the fate of -te tea, and therefore in its
modern use we must apprehend it as containing the conjunctive
form -te.
The frequent omission of wa and mo in modern colloquial, as in
' -According to Yoshizaw.i the two dots now twd for indicating syllitdes with
voiced consonants arc not more tlian 500 years old (cf. }Tr.. p. SH).
- This is in accordance with the Nihon Koten Zen-.hu version (p. 1.")). Arcordinc to
the K6( hi'i Nihon Bungaku Taikei edition (vol. xiii, p. 442), i/oti-'dule stands alone with-
out being followed by u-a. This supports my conjecture that the adjective contained
the conjunctive form -Ip, not the imperfect form.
THE JAPANESE PARTICLES WA, GA, AND MO
47
Itte ikenai (instead of Itte wa ikenai) “ You mustn't go " and Itte ii
(instead of Itte mo ii) “ You may go ”, gives us a hint that the
suppositional meaning “if” is inherent in the -te form quite
independent of the function of tea and mo. Both wa and ba used in
a suppositional (or conditional) context serves to draw the attention of
the supposition (or condition) which the speaker has just offered and
on which his main statement depends. In other words, a suppositional
or conditional construction in Japanese conveys the idea '' Suppose
(or Let it he granted) that . . . And now listen to what I'm going to
tell you ! ” It is this latter half of the concept that is expressed by
wa or ha. Similarly mo is here used with its fundamental function of
extending the application of the concept denoted by the word or group
of words which immediately follows it to the concept designated by
the preceding element. Thus Takakute mo kaiinasu means " I'll buy
it if it is cheap, as you may expect. And if it is expensive I'll buy it
all the same ” or “ Suppose it is expensive. I'll still buy it ”, hence
“ Even if it is expensive, I'll buy it ”.
The above investigation is anj-thing but thorough. Nevertheless,
it will have shown the inappropriateness of the expositions given of
the three particles tea. ga, and )ho by various grammarians. ir«
may be used with any member of a sentence or even at the end of a
clause, when it may assume the form oiba. In my opinion wa designates
that part of utterance which the speaker thinks should be apparent
or be made known to the hearer before he can give or seek information
concerning something. Thus the particle wa usually precedes that
element of a sentence which imparts or seeks information. Naturally
it can never be affixed to an interrogatory word.* When used at the
end of a sentence, as in Atashi iku wa “ I'm jolly well going ", the
particle seems to act as an emphatic reminder. Ga indicates the
relatively less known element of a sentence, restricting the applica-
tion of the concept denoted by the word or group of words which follows
it to the concept specified by the element that immediately precedes
the particle. Thus the element to which ga is affixed generally gives
or seeks information. When used at the end of a clause or sentence the
* In a sonterK'p like Sani va <10 7iarhiia'>hila “What has become of what-d've-
call-it ? ” the word iinni represents “ T»hat-d’ye-eaU-it " and is therefore not inter-
roeatory. In ancient Japanese Hvi was often aflixed to interrogatory word.s. as in
lk(i</a ua sri im " Why, you silly, how' could I say such a thing to him “ (GM.,
p. 61 ; WT., p. 70). Wa is here used after ikrig<i " how- ” to enipha.size the meaning of
the following element sa wa wdso/i “ could I say such a thing ”. Thi.s is how con-
structions of this kind generally aicjuire a rhetorical signification.
48
S. YOSHITAKE —
function of ga is reduced to one of correlating two clauses or indicating
some mental reservation. Mo mav be used with any member of a
sentence or even at the end of a clause. In each case it serves to extend
the application of the concept denoted by the word or group of words
which follows it to the concept designated by the word or group of
words which directly precedes it. This implies that the concept
indicated by the word or group of words which follows mo is shared
by some other concept than the one that is designated by the word
or group of words which immediately precedes the particle (cf.,
however, p. 43). Since im performs a duty that is exact opposite to
those of ga and mo (cf. the three examples given on p. 43), a conjoint
use of either of the latter two particles with wa would cancel each
other's function. It is for this reason that ga and mo, unlike other
particles such as wo, ni, yori, ye, made, lara, and Jcoso, can never be
followed immediately by wa.
Space does not allow me to relate how I have arrived at these
conclusions. I hope to discuss the .structure of spoken Japanese in a
further monograph in continuation of the present article.
ABBREVIATIONS USED
CH. ; B. H. Chamberlain, Handbook of CoUoquial Japan.e$e, 1907.
CK. : B. H. Chamberlain, Translation of “ Ko-ji-ki”, or “Records of Ancient
Matters ”, 1932.
FR. : K. Florenz, “ Ancient Japanese Rituals,” Transactions of the Asiatic Society
of Japan, vol. xxvii, part, i, 1899.
G. : Gikeiki (c. a.d. 1400). Text according to the Kochu Nihon Bungaku Taikei,
vol. xiii, 1926.
GM. : Genji Monogatnri (early eleventh century). Text according to the Kochu
Nihon Bungaku Taikei, vol. vi, 1926.
IK. : F. Ifa, Koryukyii, 1922.
K. ; Kojiki (a.d. 712). Text according to the Nihon Koten Zenshu, 1928.
KK. : M. Kiycda, Koto Kogoho Kogi, 1931.
MC. : W. M. McGovern, Colloqnial Japanese, 1920.
MG. : S. Jlatsuoka, Xihon Oengogatu, 1928.
MK. : Masukaganii (r. a.d. 1350). Text according to the Kochu Nihon Bungaku
Taikei, vol. xii, 1926.
MS. : Mnkura-no-Sosh i (early eleventh century). Text according to the Kochu
Nihon Bungaku T.aikei. vol. iii, 192.5.
NT. ; C. Nos.s, .4 Text-Book of Colloquial Japanese (based on R. Lange’s Lehrbuch
der japanischen XTmgang.ssprache), 1907.
OK. : Okagami {r. a.d. 1120). Text according to the Kochu Nihon Bungaku Taikei,
vol. xii, 1926.
OM. ; H. Omadaka, Man-yo-shu Shinshahi, vol. i, 1931.
ON. ; Oharai-no-Korito (probably older than a.d. 600). Text according to the Zoho
Motoori Norinaga Zenshu, vol. v, 1926.
RC. : A. Rose-Innes, Conversational Japanese for Beginners, 1924.
THE JAPANESE PARTICLES WA, GA, AND MO
49
S. :
SA. :
SH. :
SS. :
T. :
TK. :
U. :
WT. :
YA. :
YB.
YG.
YK.
YM.
YT.
YY. ;
Semmyo (Imperial edicts dating from a.d. 697-789). Text according to the
Zoho Motoori Xorinaga Zenshu, vol. v, 1926.
I. Shimmura, Amakusabon Isoho Monogatari (a.d. 1593), 1928.
G. B. Sansom. A/i Historical Grammar of Japanese, 1928.
E. A. Sonnenschein, The, Soul of Grammar, 1927.
Taketori ilonogatari (ninth century). Text according to the Kochu Nihon
Bungaku Taikei, vol. ii, 1927.
U. Tsugita, Kojiki Shinko, 1925.
Uji Shui Monogatari (c. a.d. 1240). Text according to the Kochii Nihon
Bungaku Taikei, vol. x, 1926.
A. D. lYaley, The Tale of Genji, vol. i, 1925.
S. Yoshitake, “ An Analytical Study of the Conjugations of Japane.'-e Yerbs
and Adjectives,” Bulletin of the School of Oriental Shidies, vol. vi. part 3, 1931.
T. Yamada, Xihon Bumpo Kogi, 1924.
Y'. Y’oshizawa, Kokugoshi Gaisetsu, 1931.
T. Y^amada, Xihon Kogoho Kogi, 1924.
K. Y'uasa, M uromachi-jidai no Gengo Kenkyu, 1929.
Y". Y'oshizawa, “ Iwayuru ^oo ni tsuzuru joji ga ni tsuite,” Kanazawa Hakushi
Kanieki Kinen Toybgogaku no Kenkyu, 1932.
T. Yamada, Xihon Bumpo Yoron (Iwanami Koza Nihon Bungaku), 1931.
VOL. VIII,
PART 1
4
Some Corrections and Critical Remarks on Dr. Johan
van Manen’s Contribution to the Bibliography of
Tibet ^
By Andrew Vostrikov
ri^HE bibliography of the hterature of Tibet is beset with extra-
ordinary difficulties. The access to the literary treasures of
this country is not easy and our knowledge in this domain is therefore
exceedingly limited. If we exclude the two great Collections of Transla-
tions (the Kanjur and the Tanjur) which are tolerably well known
from the bibhographical standpoint,^ we must confess that the great
ocean of the original Tibetan hterature, the hterature of Tibet proper,
remains almost entirely unknown. It cannot, of course, be maintained
that this hterature has not been studied at aU. We possess some
excellent editions, translations, and investigations devoted to original
Tibetan compositions. However, compared with the enormous compass
of Tibetan works existing but unknown to us even by their titles,
these scanty pubhcations are not at all adequate to the task. More-
over, the choice of Tibetan works for investigation and the perspective
under which they are regarded sometimes clearly show that the
* yote by Professor Th. Strherbatsky. — Dr. Andrew Vostrikov has executed several
tours through the Buddhist monasteries of Buriat -Mongolia in search of unknown
Tibetan MSS. and block-prints. His collections constitute an important addition
to the Tibetan fund preserved in the Oriental Institution (formerly the Asiatic Museum)
of the Academy of Sciences of the U.S.S.R. During his tours he acquired some
experience in cataloguing the new findings. This is not always an easy task since the
real titles of the works are verv’ often disguised either under long phrases of artistic
comjwsition or under conventional abridged nicknames.
Since the present juncture seems to offer many more facilities for collecting Tibetan
literature than the former times of the almost absolute seclusion of that country, and
si-'ce the Asiatic Society of Bengal always made efforts in that direction, the following
critical remarks by Dr. A. Vostrikov on Dr. J. van Manen’s Contribution to the
bibliography of Tibet will, I hope, not be found superfluous and are here presented to
the English reader.
® The most prominent works in this line are; “Index du Bstan-hgyur,’’ par P. Cordier
(Catalogue du fonds tibetain de la Bibliothrque yationale, ii and iii), Paris,
1909-1915 ; Yerzeichnis der tihetischen Handsehriflen der Koniglichen Bibliothek zu
Berlin, von Dr. Hermann Eeckh. 1 Abt. Kanjur (Bkah-hgyur), Berlin, 1914 ; “ Analysis
of Kanjur and Tanjur,” by Alexander Csoma Kordsi (in the Asiatic Researches,
vol. XX, Calcutta, 1836), and its emendated translation by L. Peer (in the Annates du
Musee Guimet, vol. ii, Paris, 1881) ; the works of I. J. Schmidt, A. Schiefner,
B. Laufer, and many others.
A considerable number of Tanjur texts has been edited and translated by
Professor Th. Stcherbatsky, M. de la Vallee Poussin, and others.
52
AXDEEW VOSTRIKOV —
author has made his choice at random and has absolutely no know-
ledge of the vast dimensions of the respective domain of Tibetan
literature and of the place which the object of his choice occupies
in it. It thus happens that much labour is sometimes spent on composi-
tions which are not at all worthy of such attention. ^
In order to guide the investigator and to guarantee some system
in his work, a review of all the extant riches of Tibetan hterature is
absolutely indispensable.
It is clear that inasmuch as the possibility of studying the literature
of Tibet in situ at present seems to be excluded, the work of its biblio-
graphy can only be achieved by a careful descriptior of all Tibetan
manuscripts and block-prints already existing in different European
collections. They, of course, are not complete, but in order to supple-
ment them it must be well known what they already contain. Only
then we will be able to start on the work of supplementing them, and
there is no better method of doing it than the organization of tours
in search of Tibetan MSS. and block-prints in Tibet, Mongolia, and
China, conducted on the same lines as the celebrated tours in search
of Sanskrit MSS. by Professor G. Biihler, Sir R. G. Bhandarkar,
and others whose reports have laid the foundation of a
systematic bibliography of Sanskrit literature.
^ As an example of much labour spent on an unimportant subject, we can quote
the work of Professor J. Bacot : Une g^raramaire tibetaine du tibetain classique ;
les Slokas grammaticaux de Thon.mi.sam.bho.ta avec leur commentaires, trad,
du tibetain et annotes par J, Bacot ” {Annales du Musee Guimet, BibliotMqut d^Etude,
t. 37), Paris, 1928.
It is a splendidly executed work, but the choice of the text is strange. Its main
part is not the short tract of Thon.mi.sam.bho.ta, which has been published several
times before, but its commentary, which is here reproduced with great technical skill
phototypicaUy, a transcription in Tibetan and Roman characters, as well as a French
translation being added. But just this commentary is not at all worthy of the great
attention devoted to it. It is a very popular primer ; hundreds of such elementary
manuals exist in Tibet. It represents a short and anonymous extract from the very
well-known grammar of Situ (^* H) and from the notes on it by
Considering that there are most important, fundamental grammars written by
CO -*■
Tibetans, as e.g. the grammar of which was followed by
a whole school of grammarians, a school which is constantly referred to and criticized
by Situ ; or the grammar of containing a very interesting.
quite new form of exposition and many other important and interesting works on
grammar by the Tibetans, all of them unpublished and hardly known by name ;
consifiering all this, it i.-? very strange to see Professor J. Bacot spending so much work
on a quite insigniticant tract, without at all mentioning and apparently without
suspecting the existence of works of much greater importance.
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
53
Unfortunately the work of cataloguing the original Tibetan
compositions contained in European libraries has not yet been
seriously started. We possess only several lists, incomplete and far
from being precise, mostlj’ simple enumerations compiled without
any systematic order. ^ They hardly can be regarded as scientific
descriptions. It is true that Profes.sor F. W. Thomas has made an
attempt of bringing to a systematic order the materials disseminated
in different lists in his Memorandum presented to the Govermnent of
India. This Memorandum, dated 18th January, 1901, contains a
review of all published catalogues of Tibetan MSS. and block-prints
extant in the libraries of Europe. It is a very interesting and unique
first attempt to simmiarize all that is contained in different catalogues
of Tibetan works on h'story, chronicles of monasteries, biographies,
folk-lore, and profane literature. But it is to be regretted that this
very meritorious work labours from the same defects that are contained
in the catalogues from which it draws its information — it repeats all
the mistakes contained in them.^
The bibliography of the literature of Tibet could, of course, derive
^ These lists are indicated in Professor F. W. Thomas’ Memorandum and in
Dr. van Manen's paper. Their number is now increased by a catalogue of the Schilling
collection compiled by Professor J. Bacot. Cf. “ La collection tibetaine Schilling von
Canstadt a la Bibliotheque de Tlnstitut,’* par Jacques Bacot {Journal Asiaiiqui,
tome cev, Octobre-Decembre, 1924, pp. 321-348. Paris, 1924).
2 It contains e.g. the list of “ Annals of Monasteries and Priestly Successions ’’
excerpted from the Verzekhniss der TiUtischen Handsekriften und Hoizdrucke im
A-’a'afischen Museum of I. J. Schmidt and O. Bohtlingk. But this is a mistake. The
nine works mentioned in the Memorandum under this item are not at all “ Annals *.
Xo. 1 represents a supplement to the collection of sadhanas ; Xos. 2-4
point to the literature devoted to the custom of “ total fasting ” , printed
and Xos. 5-9 represent collections of hymns
recited and sung at different divine services. These collections of hymns are different
in different monasteries, and the list of Professor F. W. Thomas contains five such
collections in use in five different monasteries in Buriat-Mongolia.
Professor F. W. Thomas equally repeats the mistake or misprint of the Verzekhniss
cs cs
in spelling and interpreting the term (prh^t^d This term is
Q.
nothing else than the Russian word ;tPn}TaT, written in Tibetan letters ; the English
“ deputy ”, It was the title given at that time to the *11 (sounds in the
Buriat pronunciation “ Bandida-Khanbo ”), or the archbishop of the Buddhist
clergy in Buriat-Mongolia, Msgr. Dambadarje Zayagiin (Zayayeff), as a member of the
Buriat -Mongolian deputation in the Imperial Commission set up by the Empress
Catherine II for elaborating the Xew Code of 1766. His work ( S* 3’ ^
“) is a short MS. containing his autobiography and the history of
51
AXDKEW VOSTRIKOV
great help from a study of original Tibetan bibliographical works,
which exist in great number, many of them being compiled by
celebrated scholars. But, unfortunately, these works are quite un-
known in Europe. The same applies to the very numerous catalogues
of Tibetan and Mongohan monastic printing offices. Nevertheless,
the main source of bibliography remains the description of funds
already contained in European libraries, since only these funds can
directly be investigated, described with precision, and their description
always verified.
However, since these funds are incomplete and casually collected
and, therefore, it is not to be expected that their investigation could
give us an idea of the whole compass of Tibetan hterature, and since,
moreover, their study requires long and assiduous work on the part
of a number of hbrarians, some investigators deemed it adffisable in
order to accelerate the very slow progress of our knowledge to have
recourse to a new method of collecting bibbographical informations,
the method of information from hearsay. But it is evident that such
a method can be resorted to in bibhography only in extremis, and
it is a priori clear that its scientific results cannot be quite satis-
factory.
The paper of Dr. Johan van Manen, “ A contribution to the
Bibliography of Tibet,” ^ must be regarded as an example of such a
bibliography from hearsay, explainable only under the extraordinary
conditions of inaccessibility of a great number of Tibetan compositions
for direct investigation. Although his paper appeared in 1923 and is
very interesting as an attempt to solve an almost insurmountable
difficulty, it has till now not been critically examined and its results
have not yet been valued. The following remarks can be regarded
as a correction of some ob\dous mistakes which can be detected even
with our hmited knowledge, and will lead us, it seems to me, to a
condemnation of Dr. van Manen’s new method of bibliography.
Willing to supply information regarding Tibetan literature not
represented in our libraries. Dr. van Manen should have consulted
the foundation of the first Buddhist Monasterj' in Transbaikalia, the Zongol monasterj'
called in Tibetan called in imitation of the celebrated
Tibetan monastery of that name). It is a very interesting account, containing a Hvelv
picture of the manner in which Buddhism has spread in Buriat-Mongolia.
1 “ A contribution to the Bibliography of Tibet,” by .Johan van Manen, Journal
and Proceedings of the Asiatic Society of Bengal, New Series, vol. xviii, 1022, No. 8
pp. 445-.525 (Issued November, 1923, Calcutta).
CRITICAL REMARKS ON THE BIBLIOGRAPHA' OF TIBET
55
the bibliographical works of the Tibetans themselves. But he was
quite ignorant of their existence. He says simply “ perhaps they
exist”. However, they really exist and many of them are widely
known. Such e.g. is “ A catalogue of works of lamas belonging to
the Kadampa and Gelugpa sects”
compiled by the celebrated Longdbl-
lama born in 1719). It is an exceedingly
important and interesting work on bibliography. It contains a list
of works which are to be found incorporated in the editions of the
complete works of their respective authors and an additional list of
works which have become the fundamental school manuals
in different monastic schools in Tibet.i Much bibhographical informa-
tion can be found in other works of the same author, especially in
his Manuals on the terminology of Buddhist religion, philosophy,
and science. Extraordinarily rich in bibhographical information
are the so-called Thob-yig’s ^ compiled by different authors.
A great many works are also mentioned in the historical and bio-
graphical literature of Tibet. Very valuable is a work called “ A list
of some rare books ” compiled by the
lama Akhu Kinpoche a.d. 1803-1875),3
and many others. But this literature is unknown to Dr. van Manen.
AH the concrete knowledge of Dr. van Manen in this department
is hmited to a work mentioned in the Catalogue of Tibetan MSS.
and Xylographs in the Asiatic Museum of Academy of Science,
St. Petersburg, compiled by I. J. Schmidt and 0. Bohtlingk,* where
under Nos. 416-456 we find :
^ This bibliographical composition is entered in the complete works of this author
under No. 25 65 folios.
2 Their importance for Tibetan bibliography has already been pointed out by
Professor W. P. Wasilieff in his paper, ** Die aufden Buddhismus beziiglichen Werke der
Universitats-Bibliothek zu Kasan,” Mdanges Asiatiques, tome ii, pp. 347-386,
St. -Petersburg, 1855. {Bulletin historico-philologique de V Academie Imp. des Sciences
de St.-Petershurgj t. xi. No. 22, 23).
® This composition is entered in his complete works, published in Ganden-rabgye-ling
c\
^ Verzeichniss der tibetischen Handschriften und Holzdrucke im Asiatischen
Museum der Kaiserlichen Akademie der Wissenschaften,*' verfasst von I. J. Schmidt
und O. Bdhtlingk {Bulletin historico-philologique de V Academie Imperial des Sciences
df St.-Pefersburg^ t. iv. No. 6, 7, 8).
56
ANDREW VOSTRIKOV —
Das von dem Lekrer Schad-sgrub-mvig-tschan {sic !) verfasste Biicher-
verzeichniss, (gennant) die Fundgrube von Kleinodien. Elf Bande.'’
But this is a mistake. The eleven volumes in question contain no
“ Biicherverzeichnis ” at all. They contain the complete works
(5|g^'a,^fi)of the verywell known lama Dagpashedub(®]!»|^’n’i)?|i;"^u)
of the Chord or monastery in Amdo. The authors of the
catalogue have mistaken for a "Biicherverzeichnis” the Table of
Contents of an edition of his complete works. According to the general
use of Tibetan, Mongolian, and Buriat printing offices an edition
of complete works never has a general title page, neither for the whole
collection, nor for its separate volumes. It represents a collection of
separate works, having each its own title and pagination, connected
together only by a letter indicating the volume number. This letter
is always printed on the left side of each folio together with the page
number. But to the whole collection or, more often, to each volume
of the complete works a Table of Contents is added where the separate
works included in the volume, or in the whole collection, are indicated
together with the number of folios in each. Such Tables of Contents
bear sometimes special names. A name of this kind is the one quoted
in the catalogue of Schmidt and Bohtlingk : “
|| ” It literally means;
“ A mine of jewels. — The Hst of works composed by the Master
Dagpashedub.” Curiously enough, another copy of this same collection
is mentioned in the same catalogue once more under Nos. 312-322.
Here again it figures under the name of its Table of Contents. The
authors, however, do not translate it literally this time, but, owing
to a remark written by hand on the binding (T*
rightly determine the books as “ die von dem Hauptlehrer,
Namens Schad-dub, verfasste Schriftensammlung ”. Unfortunately,
this very obvious coincidence has escaped the attention of Dr. van
Manen. He has thus mistaken a simple Table of Contents for a work
on bibliography. The real bibliographical works of Tibetans are
unknown to him.
To the same class of Tibetan bibliographical works may be
reckoned, to a certain extent, the very numerous catalogues of
different Tibetan and Mongolian libraries and printing offices. But
Dr. van Manen mentions only two of them and supposes without
CRITICAL REMARKS OX THE BIBLIOGRAPHY OF TIBET
57
1
I
‘'S
V
any reason that “ bibliographical pubhcations in Tibet are very rare ;
booksellers’ catalogues also
Neglecting the description of Tibetan MSS. and block-prints
already collected, and being ignorant of the original Tibetan works
on bibliography, as well as of the catalogues of every monastic printing
ofhce, Dr. van Manen betook himself to the device of ordering the
compilation ad hoc of lists of all the literature which is current in
Tibet.2 He addressed himseK to a lama with whom he was acquainted
and who was returning from a pilgrimage to India. He explained
to him the European conception of bibliography and asked him to
compile and bring Muth him on his next visit to India lists of literary
works which are in vogue among Tibetans. After a lapse of six months
two such lists were in his hands. But, as was clear from the beginning,
these lists were entirely worthless for the purpose for which they were
ordered by Dr. van Manen. They did not at all satisfy the require-
ments of scientific bibliography. This Dr. van Manen himself admits.^
1 J. van Jlanen, op. cit., p. 449.
* The attempt to receive bibliographical information from Tibet by the way of
questioning the Tibetans themselves is not quite new. In the year 1901 Dr. Emil
Schlagintweit, supported by \Y. W. Rockhill and Sarat Chandra Das, applied to
the Dalai Lama himself asking him to give his assent to a compilation of lists of old
Sanskrit MSS. preserved in the libraries of Tibetan monasteries. This attempt failed. A
letter from these scholars was delivered (in 1902) by the ambassadors of the D.S.A. and
Germany to the Chinese Government, who answered that they had fomarded the
letter to Tibet. However, no answer from Tibet was received. The history of these
transactions is related m “ Bericht uber eine Adresse an den Dalai Lama in Lhasa
(1902) zur Erlangimg von Bucherverzeichnissen aus den dortigen buddhistischen
Klostern ", von Emil Schlagintweit (Abhandtungen der philosophisch-pbilologischen
Klasse der Konigl. Bayerischen Akademie der Vi’issenschaften, xx Bd., iii Abt., S.
657-674, Munchen, 1905). We find here the English and Tibetan texts of the letter
and all the correspondence referring to the incident. Dr. E. Schlagintweit had even
the intention to make use of the authority of the German Kaiser in order to impress
the Dalai Lama with the importance of bis demand. But this scheme was finally
dropped.
Dr. van Manen is quite all right when he says about thi.s attempt : " Its legacy' is
a rather amusing and prolific description of the attempt, together with all documentary
material connected with it, which Schlagintweit published in Munich in 1904, and which
is a very serious report concerning a more or less comic episode in the progress of
Tibetan studies ” (cf. op. cit., p. 446). However, this condemnation did not prevent
hi.s repeating a similar attempt " in a more unambitious and unofficial manner ”,
as he says.
Leaving alone the manner in which the transactions have been conducted by
Dr. E. Schlagintweit and his associates, it is obvious that the question regarding the
preservation in Tibet of Sanskrit works is more or less natural inasmuch as it refers
to such works which are known, be it only by name, to have existed in India. But
Dr. van Manen’s plan of getting information regarding works quite unknown was
doomed a priori to failure independently from the manner in which it was conducted.
• Op. cit., p. 448.
58
AXDREW VOSTRIKOV —
What is still less satisfactory, these lists do not even contain the
real titles of the books, hut very often only conventional nicknames
under which they are known to the lamas and in the market.
It is not even known who were the authors of the lists and on
what authority they compiled them. They probably were not very
learned men, since the lists are full of the grossest orthographical
mistakes. The lists evidently were compiled from hearsay. The
hterature mentioned in them is chosen accidentally, at random ;
the most celebrated and widely spread departments, the works which
are more in vogue than all others, are not mentioned at all.
We find in the lists some editions-of complete works
mentioned, but those that are predominantly spread and most highly
revered are not mentioned at all. Thus the complete works of
Tsongkhapa ( and both his pupils Gyaltsab
and Khaidub )
are not mentioned at all. The complete works of all the Tashilamas,
of Taranatha, and many other celebrated authors have also escaped
the attention of the anonjunous compilers of both Dr. van Manen’s
lists. Out of all the numerous works of Gyaltsab only a single one,
which is not very important, is quoted. The same applies to the works
of Taranatha, Buton etc. In the first Ust the
editions of the Depung monastery are indicated, but
only 46 items are named, while in the catalogue
as Dr. van Manen himself states, 285
titles are given.^
Out of the enormous historical literature ^ only five works are
mentioned, and such leading works as the history of Buton ® and
^ Ibid., pp. 475-6 and 524-5.
^ Cf. the very detailed review of the historical and biographical literature of Tibet
contained in the celebrated work on the History of Buddhism in Amdo, by
(born 1800). This history was
■N
written in 1833 and is known under the abridged title The full
title is ;
3 vols. The bibliographical review is found in the first
volume, f. 4 recto, 1. 6 - f. 19 recto, 1. 4.
3 Its full title is
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
59
not to speak of many others, which are very much
read, are not mentioned at all. The same refers to the literature on
medicine and astronomy, which is also very rich. I myself have
come across a special list of Tibetan works on medicine compiled in
a Mongolian monastery. It contained 298 items. But in Dr. van
Manen's list only very few medical and astronomical books are
mentioned.
The exceedingly vast Uterature of school books is almost totally
ignored.^ Some classes, as e.g.
are omitted. Some others are represented only by single and not by
the most important manuals.
Under these conditions it is evident that the lists so compiled
are deprived of every value and it is difficult to understand
how Dr. van Manen could suppose that “ they give us the picture
of the literature affected by the modern Tibetan intelUgentzia
(sic/) ”.®
Having thus exaggerated the importance of the lists he had received;
Dr. van Manen was confronted with the very difficult task of inter-
preting or even guessing the meaning of the conventional nicknames,
of which they were full. Such an interpretation required great bibUo-
graphical knowledge, which was not at the command of Dr. van Manen.
And, therefore, his commentary is very often quite faulty.
In the sequel we will add some critical remarks on particular
composed in 1322 by (1290-
C\
1364). The work is edited in the monastery and includes 244
oblong folios. It, strictly speaking, represents a systematic review of the whole canon-
ical literature of Xorthern Buddhism translated into Tibetan, but the review is supple-
mented by an introductorj- historical sketch of the evolution of the Buddhist doctrine
in India and its first steps in Tibet. This review and this sketch are translated by
E. E. Obermiller in the Materialien zur Kunde dea Buddhismus, herausgegeben von
Dr. M. Wallcser, Heft 18 and 19, Heidelberg, 1931-2, with an introduction by Professor
Th. Stcherbatsky. The remaining part of Butdn’s work, the part containing the
Catalogue of the Tibetan Buddhist Canon, is not yet issued.
1 It was composed in 1476-8 by the celebrated translator Shonnupal, native of
Goi, (1392-1481). It was edited in the
•y- -y- C\ ^
monastery in Tibet and later on in the monastery in
Amdo. The second edition consists of two volumes and includes 568 oblong folios.
c\
^ An idea of the extent of the obligatory school manuals only, can be
gathered from the above-mentioned bibliographical work of Longdol-lama
® J. van Manen, op. cit.. p. 448.
60
AXDREW VOSTRIKOV —
items of Dr. van Manen's Index, which contain especially misleading
indications.
The lists of Dr. van Manen do not follow any systematic order.
In his index to them he follows an alphabetical order, independently
from the circumstance whether the first letter of the title refers to
the author's name, or to the title of work, or even to the name of
the country where the author was born or flourished.
We, therefore, in our notes do not follow quite exactly this un-
systematic order, but begin by pointing to the most conspicuous
mi.stakes.i
“28. (188). Explained as the 'easy introduc-
tion ’ of the college {or house) of the old pulpit {or seat). iSo further details,
philosophy, Gelulcpa."
The e.xplanation is wrong. It is clear that Dr. van Manen knows
neither the meaning of nor the meaning of In the
sequel he gives another interpretation of the last term, as “ the book
for the use of the body (assembly) of novices
is the name of a course of Eristics taught during the first
three years to novices in the monastic philosophic school *
or It is divided in three degrees according to the degree of
difficulty. Its aim is to train a no\dce in the art of right and fluent
argumentation by teaching him how to put every argument or
even every idea into the form of a regular syllogism ; and also to
teach him quickly and rightly to react on every counter-argument
set out by the opponent. The opponent questions, the disputant
answers. He answers briefly in one of the four ways. (1) If he agrees
he says simply ; “ Yes ! ” {'^f^’). (2) If he doubts, he asks ; “ For
what reason ? ’’ (^^’5^*), if the reason is not given. If he disagrees
he must at once detect where the fallacy lies, and two possibilities
are open : the mistake lies either in the minor
or in the major premise = ^nfR)- (3) If the fault lies in the
minor premise he says: "The logical rea.son unreal”
i.e. not contained in the subject, or minor term. (4) If it lies in the
^ In our quotations the figures before the Tibetan titles refer to Dr. van Manen's
Index, the other figures (in brackets), after the titles, refer to both hi.s book-lists.
2 J. van Manen, op. cit.. Index, Xo. 120.
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
61
major premise lie must at once answer : “ No invariable connection
between the middle term (or reason) and the major term.
Having this aim in view the course of begins by some
examples of discussions on the connotation of the simplest notions,
e.g. colour in general and particular colours, the form in general and
particular forms, etc. A double result is achieved, the novice gets
quite precise logical definitions of current notions with which logic
or philosophy are dealing, and he acquires a proficiency in applying
the syllogistic formulation (with its three terms) to everv argument
at the same time the author of the first school book on it, according
to Tibetan tradition, is the Tibetan lama who
lived in the twelfth century (1109-1169).'
The hterature on is enormous, it is not possible to review
it here. Two works only are indicated in Dr. van Manen's lists, but
not the most important ones. They are and
.2 I_,ea\’ing alone the second of them, whose author is the
celebrated Desi, or Regent of Tibet, Sanggye-gyamtso
and considering the first, I must repeat that Dr. van Manen
has not at all succeeded in grasping the meaning of This
term means “ the old abbot ” ; it is a nickname of the author of the
work. His real name is He was the pupil and
collaborator of the first Jamyangshepa, NgagM'ang-tsondiii («,£.»)•
The title of “the old abbot ” was
given to him as a reward for his very long tenure of office as abbot
of the Labrang monastery in Amdo. He has composed two works
on the system. The title of the first is :
fi better known under the abridged title:
f'
'■-P; ^ The Table of Contents of this work is given by Longdol-lama in his Terminology
% of Logic
4 entered in his complete works under No. 14 (^), 27 folios), f. 2, verso, 11. 2-4.
■% * J. van Manen, op. cit.. Index, No. 28 and 120.
5 ^ Cf. below, our remark on No. 121 of Dr. van Manen's Index.
62
ANDREW VOSTRIKOV
It was edited several times, in Labrang in 1860 and later in the
Aga monastery ^ in Buriat-Mongolia.
His second work is mostly known under the title
or or It was never
finished and has also never been given a title. In the Amdo edition
it has neither title nor colophon. In the Aga edition it has been given
the conventional title of
>o
or “ An unfinished composed by the venerable
Master for the use of the lama-incamate
for the benefit of his intellect ”. It contains 130 folios. In the
catalogue of the printing office of the Aga monastery ^ its name is
given as or “Unfinished
known (according to tradition) as a work compiled by the old
abbot
It is impossible to determine quite exactly to which of the two
books Dr. van Manen’s mention refers, since both can be designated
by the title But it is more probable that it refers
to the second, unfinished work, since it is more usual to quote it under
this nickname.
“ 35. (32). Commentary on the feace of the
true number{s). philosophy.” (I !)
“ 36. (4)- The method to arrive at the true
number{s). philosophy. Gelukpa.” (? !)
’ The Tibetan name of this monastery is
2 The title of this catalogue of the printing office of Aga monastery (in Buriat-
Mongolia), which we will have several times the opportunity to quote, is
'V'
It is divided into four parts, according to the length of the folios ; (1) “ catalogue
of the long blocks ” ; (2) catalogue of the “ intermediate ” blocks
(n3^’f<5|q*);
(3) catalogue of the “short” §[;,•) and (4) “very short”
Mocks.
Cf. Aga catalogue f 0 1. 3.
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
63
Here orthographical mistakes on the part of the author of the
lists have utterly changed the meaning. “ number ’’ is a mistake
for “indirect meaning”, “peace” is a mistake for
“ground”, and “method of reductio ad ahsurdum” is a
mistake for “detailed investigation The first title
accordingly means “ a groimd commentary on the division of the
meaning of Buddha’s words into a direct and an indirect meaning
The second should be translated “ a detailed investigation of the direct
and indirect meaning (of Buddha’s words) ”. These titles refer to a
very well known and widely spread class of Tibetan hterary works.
It is astonishing that neither Dr. van Manen nor his informant have
recognized it. The orthographical mistakes could not have prevented
identification, for they are very common.
• qn • is the designation of a whole branch of Tibetan literature,
containing numerous works. They are all devoted to the same problem
of a division ) of Buddhist scriptural works into two classes,
the class of direct meaning, called or (corresponding to
Sanskrit and the class of indirect meaning, a meaning adapted
to the comprehension by the auditory, called or
(corresponding to Sanskrit As a matter of fact, Buddhism
during its long history many times altered its fundamental philo-
sophical and religious views. It therefore became a necessity for
Buddhist authors to explain the connection of the new ideas with the
old ones, to explain their contradiction and to justify the change.
It was impossible simply to reject the old ideas and to replace them
by the new ones. It was impossible especially in those cases when
a firmly established tradition believed that the meaning in question
was established by Buddha himself. Such a meaning could never be
represented as heretical. But its glaring contradiction with the new
ideas could totally undermine the canonical authority of the latter.
The solution of such difficulties which are not unknown in the
history of many religious systems lies always in the creation of a new
branch of literature whose aim it is to clear off the most flagrant
contradictions by doing violence to the original text and its plain
meaning. The original text then becomes so interpreted that the
contradiction disappears. In Buddhism we have two great innovators.
^ Cf. below our remark on Xo. 141 of Dr. van Manen s Index.
6t
ANDREW VOSTRIKOV —
the Masters Najjarjuna and Asaiiga. the founders of Mahayanad
In Tibet the celebrated learned monk Tsongkhapa, whose writings
have served as the ideological platform for the reorganization of the
Buddhist church in this country, has written a special work dealing
with this problem of interpretation. Its full title is
It is better known
under the abridged titles and
or under the short, conventional nickname
This highly esteemed work of Tsonkhapa, who in Tibet has been
given the title of a second Buddha, is founded on a comparison
and interpretation of two canonical works :
and
Its aim is to demonstrate the method to be followed in order to clear
away “supposed’' contradictions, “supposed” from the author’s
point of view . This work exists in innumerable editions and has been
commented upon by a great many authors. In the monastic schools of
Tibet and Mongolia it forms a special course, as a part of the
(m'^t^rfrr) class. For the use of these courses special manuals
have many times been composed. Two such manuals are mentioned in
the lists of Dr. van Manen.
The first of them, is known to me only from
hearsay. Its author is supposed to be a learned Mongolian lama of
the name ofNaiman Toin (
The name of the second work is an abridged
title appertaining to a whole class of similar works. It is, of course,
impossible from such a generic title to determine exactly what work
is meant. But since it i.s mentioned in connection with other school
manuals in use in the school of the Depung monastery.
it becomes most probable that the work
1 In Tibetan they are called the two Great Vehicles of
Mahayana.
2 Cf. Kanjur, Mdo, voL v (^*).
^ Ibid., Mdo, vol. xvi (^').
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
65
n3:,’f5,gy ” is meant. Its author is a
u I
(1478-1554). It contains 59 folios and is one of the
chief manuals («=!]'*) of that school.*
“56. (169). Cmnmentary on the obstructions
of the Tantrilc Lamm ? (Very doubtful, but so according to oral
information.y'
The knowledge of the author of this piece of information could
not be very extensive, for the title really means ; “ The commentary
by Asahga (^^"jii;) on (Maitreya's work called) the Sublime
Science = ■^'W^fT^).” The abridged title
is the current name under which this most popular and widely
spread work, one of the fundamental canonical works of Northern
Buddhism, is very well known to every Buddhist monk. The
of Maitreya and the commentary on it by Asanga, called
oi ti ^ l, are incorporated (in Tibetan translations) in the Tanjur.® and
there are besides innumerable editions in Tibet, Mongolia, China, and
Buriat-Mongolia.® Both texts are edited in Aga monastery, and
in the catalogue of the printing office of this monastery the
Commentary of Asanga bears just the same abbreviated title as the
one found in Dr. van Manems list, viz. 78 fohos.*
CN
^ Longdol-laraa mentions this work among the obligatory manuals of
the school of the Depung monastery. Compare his bibliographical
work indicated above, f. 44, verso.
This work is also contained among other manuals used in this school, which were
purchased by the late G. Zybikoff in Tibet. Cf. Musei Asiatici Pelropolitani Xoticiae,
iv. CniicoK TiioeTCKiiM iisaamia.M, npiiBe.ieniiHM IIuOiikobum b 1902 r,, Xo. 21
(St. Petersburg, 1904). The Tibetan title there given
is not correct. ZybikofF's collection contains no complete works "
of but a series of obligatory manuals (unfortunately
•V“
incomplete) of the school, a series containing, among others, several
works of this author.
2 Cf. Tanjur, Mdo, vol. xliv Cf. Cordier, op. cit., p. iii, p. 374.
2 Both texts have been translated into English by E. E. Obermiller. Cf. Acta
Orientah'a, vol. ix. pp. ii-iii, 1931.
CN C\
* Cf. the Aga catalogue quoted above, , f. 2, verso, 1. 1.
VOL. VIII. PART 1.
66
ANDREW VOSTRIKOV
The title 5^'? = altogether unknown to Dr. van
ilanen, since we find in his Index the following interpretation :
•‘201. (1-*^)-
account of the Lnina succession (doctrinal tradition) of Translator
Lodansherab. 5s‘5' = 5’ 5^" ■S’ ='
^ '<S
The interpretation is ^vrong, since the title really means : “An
abridged exposition of the sense = fRTjgr^d of Lttara-
tantra composed by Translator Loden.sherab (=
A.D. 1059-1109).'
“67. (54). The icorhs of iSgahcang
Jampa . yamthar. Said to have been the teacher of the present Dalai
Lama and to have died some 20 gears ago at an advanced age.
‘words spoken big’.'’
Again a quite fantastic determination. According to Dr. van
Manen's list the or Complete works of
Ngagwangjampa consist of three volumes, and their blocks are found
in the printing office of the Purchog mona.stery
If Dr. van Manen would have given himself the trouble to look into
G. Zybikoff's Jovnicg to Tibet.- he would have known that in the
monasterv blocks of the complete works of one of
the former lamas-incarnate Xgagwang-jampa are pre.scrved. in three
volumes. The general name of all the lamas-incarnate of the monastery
of is and ZybikofF. who purchased
there a copv of the thri'c volumes of his works, quotes them under
that name.^ Longdbl-lama in his bibliogra^ducal work, quoted above,
mentions all the works of this author contained in throe volumes of
his ■'Complete works".-* They exactly correspond to those found
in Zvbiko.T's collectioTi.
Dr. van Manen simply confounded two lamas-incarnati- of that
monasterv. who both bore the title of Imt of
1 ,T. v.iti Manen. op. cit., p. 47S.
- r. ..rj.'.i.i'it'T iiii.ioMMiii: y cnHTi.nih 'rnoeTii jfo ;pi*‘BinncaM iumch-
liLiM D r.r.‘' PyccKom reorp.MlniqpcKoro nOmecTRa. Ilciporpa^,
1019). p. 3.5-2.
Cf. Xr>(i<'ittr, IV. ]), 3. No. 23-
* Cf. hi< bihliu^'raphical woik quoted jil)ovo, f. .34, info. I. 2 f. .3.5, rrcfo, 1. 2.
CRITICAL REMARKS OX THE BIBLIOGRAPHY OF TIBET
67
them the one was called ; he is known to Longdol-
laraa, and is the author of the three volumes indicated in Dr. van
Manen's list and in Zybikoff’s catalogue. The other, called
, was indeed the teacher of the thirteenth
Dalai Lama, died verv old in August. 1901.' but he has nothitiE to do
with the three volumes of the works of the first. Both are considered
and revered as two incarnations of the same person, but that is not
enough to make them identical.
“83. (182). The two
nernsions, full and medium, of (the worh) conducting towards God
Mahjiighosa, bg (Lama) Gala {Jogmind uncertain, see S.C.D..^
■s.r. but see next number. There is also a (=3‘)."
I < s
“84. (164). About the writings
of Gala concerning Mahjughusa. Uncertain. See previous number.
Sandhar."
Here the mi.stake apparently is due to Dr. van Manen himself.
cs . C*
He has confounded the words 'a* (5^’ with =l’|Sq‘. It is rather difficult
q ' q '
to distinguish them in cursive wTiting. In his interpretation of these
items Dr. van iManen has. moreover, disrupted the name of the
author. He has misunderstood that the e.xpression
is nothing but an abbreviation of the author's name —
o, = q • g - q • q a, • qo. . ^ _
whose name i.s so curiously disrupted
by Dr. van Manen, was the third Abbot of the Depung
monastery; in this capacity he is also known under
the name of a,Eq’ Yq^q''q!:|q-a-?Fq-'q3;-.3 He lived a.d. 1429-1503.
means '■ The detailed and
the intermediate introductions into the (iMadhyamika) .system,
(the works) of Jamvang-Gabi-Lodoi ", and reftu'.s to two of
his three verv celelirated works devoted to an Introduction into the
Madhyamika system, the more detailed
' Cf. r. If. Ubi'iiiKon. op. fit., p. 270.
" i.e. .4 Tihrfan-Efttflish Du'tinmiry. mmpilcd by Sarat Chandra Da'«
^ Cf. the above-mentioned biblio^ra]>h!ial work of Longdol lama, f. 44. versOy
I!. 1-2.
68
ANDREW VOSTRIKOV
17 folios, and the intermediate
12 folios. His tliird work, the “abridged”
redaction is not mentioned.^
means “A collection of Jamyang-
Gabi-Lodoi's works”, means, .sometimes, an incomplete
collection, distinct from “Complete works".
“ 92. (167). Discourse { cE^i]^’) on the occasion
(^^•) of the exhibition of the scroll (or banner, on the Sundays.
Refers to a custom in both Lhasa and Tashilhiinpo to exhibit neicly
painted thankas on Sundays, to the Dalai and Tashi Lamas.'"
It is difficult to imagine an interpretation more misleading.
Dr. van Manen evidently does not know that the works on Yinaya
contain a special chapter or a supplement devoted to time-
computation and the calendar ; the system by which beginning,
length, and subdivision of a civil year is fixed ; discussions on the
seasons, their number ; on months, festivals, and expiatory days.
This supplement is called or . A whole
literature is devoted to this subject in Tibet.^ A special course on
time-computation ) is gone through in the mona.stic schools
as a part of the One of the best known and widely
spread school books on this subject is the work contained in Dr. van
Manen’s fist under the title
The first word (or is the name of a place and of its
monastery, one of the many mona.steries belonging to the so-called
P^rt of the abbey.® But in the present case is
an abbreviation for or y«*gc;*n’, the nickname of the author
* These three work.s are mentioned in the liihliowraphieal work of Longdol-lam a,
f. 44, rpr.' 0 , 1. 4. They arc contained in Zybikoff's eolleetion. No. 21. vol. i.v.
- A short summary of this subject i.s given in the work of Longdoldama :
f. 22 , cers-o, i. 2
sqq. This work is entered in the complete works of this author, under No. 12 (^’).
3 Cf. the work of Longdol-lama, called
-y-' 00 O.
, f. 4, nctOf I. 4. This work is entered in his complete works,
under No. 22 (S').
CRITICAL REMARKS OX THE BIBLIOGRAPHY OF TIBET
69
of this book, the lama of monastery. His full name is
y j) • g:;* n • n • .
is his work. Its full title is
library of Oriental Institution of the Academy of Sciences, U.S.S.R.,
possesses a copy of its block-print and I myself have another one.
printed at the Aga monastery in Buriat-Mongolia ; this block-print
includes 71 folios. In the catalogue of the printing office of the Aga
monastery it has the title; “
“101. (127). Text and
Commentary of the ‘ Omniscience ’ of Talishang Lotsava. Talctshang
a place in Tibet, the birth-place of the Lotsava.’"
Not knowing neither the text in question, nor its real title. Dr. van
Manen had no oth^r issue than to translate literally its abridged
title which figures in his list. This method followed by him in almost
all his notes inevitably leads to mistakes. The real title of the main
text is:
; edition of the monastery in Tibet.
11 folios. The title of its commentary is :
edition of the same monastery, 33 folios. As is seen from the title,
this text is devoted to a consideration of all “ the five sciences ”,
i.e. esoteric philosophy (3jC,'^q’n), grammar logic
a • ), technique (q^'y^i'n'), and medicine
M'ith the unique aim of establishing the cardinal tenet of all Buddhism,
the denial of the category of substance from every possible point of
view. The author of both these works is a very well-known heresiarch
bom in 1405.^
* Cf. Aga catalogue, , t 6, verso, 1. 6.
^ About his heresy as well as the heresy of ^*‘^^*^* it is told that neither in
India nor in Tibet is there anything similar to them. Cf.
History of the Rise, Progress, and Downfall of Buddhism in India and History of Tibet
from Early Times to 1745 a.d., by ^»1 Ed. by
Sarat Chandra Das, Calcutta, 1908, p. 197.
70
ANDRFAV VOSTRIKOV
“38. (128). Commentary and text of
the Diibtha Kiinshe, the Siddhanta leading to omniscience ”
Here the same mistake. The full title of the work is ;
edited in the
monastery I'f fohos. The title of the
commentary is
edited in the same place,
114 folios. As is seen from the title and the text of these works, they
are devoted to a consideration of all philo.sophical systems in their
bearin" upon the problem of infinity. The author of them is the just
mentioned I possess a copy of an
edition of these two works together with the two works of the same
author mentioned above.^ They make together a small volume in
which the four works are noted by letters but each has its
own pagination. •
“ 121. (202). Commentary on the
five, ways of composition, by the Desi,
co>n-
position.”
Again a quite fantastic interpretation. It is a work on Sanskrit
grammar, a commentary on the very well-knmiTi
is a literal translation of the Sanskrit word
a designation of the changes of final sounds under the influence of
a close contact with the initial sound of the next word. Here the
five first chapters of the are
understood : the first chapter on sound.s in general, the second on
sandhi of vowels, the third on exceptions, the fourth on sandhi of
consonants, and the fifth on visarga-sandhi. All five chapters together
represent the part on phonetics of this grammar and are known under
the name -.s
^ See previous remark.
“ The Tibetan translatum of the
is fnunT in Tanjur, Mdo, vol. exxxiv
^ Cf. e.?. the special work nii this subject, loinpnsed liv tlie
Lama of Choni Its title is
folios.
above-mentioned
V
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
71
The list of Dr. van ilaiien very clearly points to the beginning of
an unfinished commentary on the Its full title
Its author is the celebrated Desi. i.e. llegent of Tibet
Sanggye-gyamt.so who ruled the country
from 1679 to 1702. and was killed in 1705. He was a prolific
author and has left an inheritance of many works on diverse
subjects.*^
141. (!)• {Rendering altogether
problematic) : Method of revering the pdramitds of the mddhgamikas.
The expression has not been explained and the grammatical
relation between and'^^'^‘ is uncertain. . philosophy."
The rendering is not altogether problematic ”. but it is altogether
^ Dr. van Manen's Inde.v mentions under Xo. 122 an edition of liis complete works,
and under the Nos. 120, 121. l.jl. 152. and 1.53 Ins different works separately. No. 120
indicates his work on Eristics (cf. above our remark on No 28 of Dr van Manen',-
Index) ; No. 121 represents his work on Phonetics: No. 151 lefers to his great
work on Astronomy : No. 152 indicates his celebrated work on Medicine ; and
No. 1.53 represents his work on the history of the Gelugpa sect.
No. 187, gUfA'ffif. refers very probably also to a composition of this author,
whose full title is
of
objections ■which were raised against another work of his. the
473 folios. It contains a repudiation
3*^*' 633 folios. Both the^e work^ are indicated in Schmidt and Bohthngk's
YfizekhnisSj under Xo^. 432 and 433, anil in Zybikoff's Catalogue under Xos. 31 and
29. The^e work>, as is clear from their titles, are devoted to Astronomy. Tlie Hrst of
them i'' that very work which A. C'^oma Korusi mentions in his Ornmuiar of the
Tihetan Liinguatje iu English (Calcutta, 1834). p. 191. and the secoml is that very
work from which he borrows his Chronological Table (ibid,, pp. 181-191).
From hearsay I myself, as well as Dr. van Manen, have heanl of the existence of
•n
a medical work ^111'^*^'^'. but even if this rumour is founded, it is hardly meant m
the present case. It is much more probable that the very well known work of this
author, the work mentioned above, is meant, since it appears together with his other
work, as an edition of that very printing ofKre, viz. which is known
as the plac'e where this astronomical work is usually printed.
72
ANDREW VOSTRIKOV
wrong, for this item clearly refers to tw'o commentaries of a special
type. One is the so-called “ detailed investigation ” of the paramita
doctrine. The other is a “ detailed investigation “ of the niadhyamika
system of philosophy.
“ The grammatical relation between and is un-
certain,” says Dr. van Manen, while it is the simplest relation of
juxtaposition where the particle ‘"and” is omitted. This is
very often done, especially in titles.
is a mistake for ; this orthographical mistake,
however, is very habitual even among learned men, because both
words sound similarly and are closely related as regards their con-
ventional, technical meaning. The term can be rendered as
“ the method of rednctio ad absurdum ” or as “ the deductive method ”
in general. It corresponds verj' closely to the technical term
For a man not knowing the exact spelling of the latter, can
appear as the most natural spelling. In the same manner the term
is sometimes confounded with the word because
both terms are very similar in pronunciation and in their meaning.
The term or can be literally rendered as
“ detailed investigation ” or “ detailed consideration ”. As a technical
term it is a designation of a special form of commentaries on funda-
mental texts which has been called forth by the necessities
of teaching in the monastic schools of Tibet.
In the original Tibetan literature of commentaries we find three
chief types: and
The first of them, represents a paraphrase, a continuous,
connected exposition of the same subject, containing almost every
word of the main text, but in a changed mode of syntactical expression
and wdth additions. It is an explanatory paraphrase of the main
text where all its verbal wealth is made use of. Having regard to the
fact that it is a habit in Tibet to consider every work, even a quite
independent one, as a commentary on some sacred text, every literary
work written as a continuous and connected exposition of its subject
matter may be called There are two varieties of the ;
(1) “the interpretation of words” and (2) “the inter-
pretation of meaning ”
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
73
Tte second type, or more precisely , is a commentary
of the usual type. It contains notes and glosses to separate passages
which require explanation. It very often happens that these glosses
are numbered in order to facilitate their being used when reading
the main text ; they correspond to our “ footnotes ’’.i
The third type is called In contradistinction to the
other types it contains a critical consideration of the main text. It
contains all its different interpretations, all right and wrong opinions
which have been expressed on it, a proof of the right ones and a disproof
of the wrong ones. By- a construction of a whole series of inter-
connected syllogisms it tries to establish that the opinion of the
opponent is contradicted by Scripture, by experience, and
by his own views. This method of deductive ratiocination is
called in Tibet (from the wmrd ;). This is
also the reason why insufficiently educated men often write
instead of
Independently of this its chief aim of a detailed proof of one’s own
and a repudiation of contrary opinions, a has a second aim,
viz. the aim of training novices in the art of conducting a debate.
Every discussion is an example to be followed in every bilateral
disputation. Therefore, in the monastic schools, namely, the philo-
sophical schools of the dominating sect, the so-called
* Cf. e.g. the work of the Mongolian lama, Ngagwang-paldiin
more known under the surname bearing the title:
3 vols. The work was composed in 1846 and printed in the Great Monastery of Ourga,
Mongoha.
This work represents a commentary on the celebrated work of the first
Jamyangshepa Xgagwang-tsondui
1648-1722 A.D.), caUed per abbreviation the fuU title being
It
was edited several times in Tibet, Ando, and Mongolia. It was also edited in the Aga
monastery in Buriat-Mongolia. It is mentioned in Dr. van Manen's Index under No. 39.
It is contained in his first list among the publications of the school of the
Depung monastery ; the obligatory manuals of that school are mostly the works of
the first Jamyangshepa.
74
ANDREW VOSTRIKOV
^*^’‘9^*^**^* the manuals are very much in
vogue ; they are among the obligatory manuals ( uj’J] ‘ 56 ’ ) of every school .
There are on every branch of learning without exception.
always presupposes the existence of a fundamental text,
“the root” (g“^* = ?j^), which is alwavs the same, but the
is different according to the school. The number of
is enormous, but the greatest part are devoted to such
subjects as and since these doctrines represent the
focus of lamaist .scholarship.
The first of the texts mentioned in Dr. van Manen’s list is
a dialectical commentary on the
The second is a commentary of the same t}'pe on Candrakirti's
(jrr^JTefiTSfrlTTO-" latter is the fundamental
text in the .study of iladhyamika system in the monastic schools of
the sect.
Tt is inipossihle to determine quite exactly who the authors of
these works wore, because we have in the list only their generic name,
the name of the class of works to which they belong. But considering
that in Dr. van iManen's list they stand at the head of all the funda-
mental manuals of the school of the Depung monastery.
it becomes very probable that they are the manuals admitted in that
school. We probably have here two works of n
who really is the author of two such works, admitted as
obligatory manuals of that school, as recorded by Longdol-
lama.^
■■ 155. p) ' (11). The jive chapters of Loverehgion .
i.e. Lama Scramonaslenj."
A very categorical, but quite fanta.stic interpretation ! It is even
quite incomprehen.sihle who cotdd have suggested such a monstrosity !
^ Cf. Tanjur, Mdo. vol. i ("]*). The Sanskrit and Tihetan texts of this work are
jadilishetl by Profe:iSor Th. Stchcrbat&ky and K. E. (ibermiller lu the Bthliofhera
Bwhlhica, xxiii.
- Cf. Tanjur, Mdo, vol. xxiii The Tibetan text of the Madhyamakavatara
is published by M. L. de la Vallee Poussin m the BihVmthrn BwJ'lhica series. vt)l. ix.
^ Cf. his bibliographical work mentioned above, f. 44. 11. 4~G.
CRITICAL REMARKS OX THE BIBLIOGRAPHY OF TIBET
75
The title is a most usual, most Modely spread,
most popular name generally given to the collection of the
celebrated five treatises of Maitreya. They are: (1) =
= ® ; (d) * ;
and (5) IVS'**’ =
Under the title of which is more usual than the
identical title these five works of Maitreya are known
in the whole domain of Tibetan literature and even to every half-
literate Tibetan.® Under this title they appear in all the catalogues
of all the printing offices of Tibetan and Mongolian Monasteries
and even have found their way into om European scientific literature.®
The quoted examples clearly show the difficulties of the task
assumed by Dr. van Manen when he resolved to publish his two
quite inadequate lists. Me have by no means pointed out all the
mistakes of which his Inde.x is full. M e have .selected only some most
1
1 Cf. Tanjur, JIdo, vol. i (T).
^ Cf. ibid., Mdo, vol. xliv ( 5 ’). Sanskrit te.xt (with comment. arv) was edited and
translated by St. Sylvain Levi in the BihlinlKquc tie VEi'nlc dei Hnutef EtuduK. Sciences
historiqiies et philohgiques, fuse. 159 et 182, Parts, 1907-1911.
3 Cf. ibid., Sido, vol. xliv ( 5 *). The Sanskrit text of the subcoimnentary
on it, composed by Sthiramati 1 '' Partly
published bv Dr. Vidhnshekhara Bhattaeharya and Professor G. Tucci in the Calcutta
Oriental .Serie.s. Xo. 21. Calcutta. 1932 (only the first part issued).
* Cf. ibid., Mdo, vol. xliv (^').
s Cf. ibid., Mdo.. vol. xliv (^’). Cf. also above our remark on Xo. .70 of Dr. van
Manen's Index
® Cf. c.c. a special work of the Choni lama 71*^ r| (born 1727),
devoted to a review of the contents of these live treatises of Maitreya. This work is
10 folios.
“ Cf. e.g. the Aga catalogue mentioned above, f. 2, rersn,
11. 2-3.
* Cf. B. IT. Bacii.ibcii. ..BymiHM.,er.) sornaTU, iifTopiiii it ,iiiTepaTvpa,'‘ part in,
p. 119, footnote 1. Cf. also B. Bapa,Tiiiiii. ..CTaxvH MaiiTpeii B Bo.iotom Xpaxie n
JIaBpaiie.“ Bibliotheca Buddhica, x.xii, p. 2.7, footnote 1. etc., etc.
76
CRITICAL REMARKS ON THE BIBLIOGRAPHY OF TIBET
conspicuous and suggestive examples. Almost every item wants
either correction or supplementary explanation. Without going into
details we can shortly mention that even such a work as Bodhicarya-
vatara which is known to every Buddhologist has been overlooked
by Dr. van Manen, when he explains the abridged title of its com-
mentary, “ as ‘'Commentary on entering on
the (right) practice (or conduct) ^ Not less wonderful is
the interpretation of the term as “ Imperial as “ an adjective
of rank, derived from the designation of the Emperor of China,” ^
whereas this term means “ superior former ”, and is applied to the
first reincarnation of a celebrated lama ; the title in question,
refers to the first Changkya-Khutuktu, Ngagwang-
chbiden who hved in 1642-1714 ; etc., etc.
It hardly is necessary to go through the whole of Dr. van Manen's
Index and point to every mistake and obscurity which is found almost
in every item. Since it is clear that the lists themselves, as mentioned
above, are unsatisfactory, it is useless to attempt their full emenda-
tion. Our aim was not to correct them, but to show the mistake
of the method of collecting bibliographical informations from hearsay.
Our review of Dr. van Manen’s attempt confirms us in the convic-
tion that Tibetan bibhography must start just by a careful description
of all Tibetan literary works to be found in the libraries of Europe
and by a study of original Tibetan compositions on bibliography
and history, as well as of catalogues of monastic printing offices
(3^’) in Tibet, China, Mongolia, and Buriat-Mongoha.
In this department very much can be done. The description of
all the e.g., would be very useful, since nobody knows in
Europe what works they include.®
Although our collections are very far from being complete, never-
theless they represent the only rehable material in our hands and
every precise and detailed description of them becomes a real and
valuable contribution towards a bibliography of Tibet.
* J. van Manen, op. cit.. Index, No. 136.
2 Ibid., No. 72.
^ In his introduction to the edition of the Sarataniintarasiddhi (Bibliotheca
Buddhka, xix) Professor Th. Steherbat.sky has given the full list of the works of
But his example has unfortunately not so far been
q ' ' '
followed by other scholars. *
BfLL. S O.S. Ycl. VIIT. Part 1 .
Plate IT.
.^3 E^» 3 •?. aV^s M»*
'-. ■ • •*>'■ . '• '*v. -
» «-3a'3'3ji -i:£^5-*^J
i ^ X •‘’‘' Z +~ ^ m‘
k- . -Kj,
?*'»:•=■'» -5 -r’
i ** . ■■ ^ ^' *■
S' s J -7-4-'Jff^f'^"V«3j''J»Ar'* ‘*'
« ■> «
j T'J-^«''i>"**i5j *~li?'*»’***'^|A
! -^4 *J’»*3) --.s
‘i . . „
• -»-f ^^-.•«»‘a^p, 'I
?;?«!-? ,54.4.3
* 5 T-.|^,r J i J? 4.,yr^ ^ ^ •
i . » ’ - J «
^ ^ I =»-3 ?£ . ^ 4 > r »^' 1
^■~» »|>. ?4e
I A4 , : , . ,y^ - ^ 3
i 4 5^ i , ^ ^ ^ ^
^ V ? “ •j> t 'T- -I^<? r a J s » »
I ■. 1,1. .
r
S\x-.M»ii \M, S\K\ INni. ('n. or IT.
A Fragment of the Uttaratantra in Sanskrit
By H. W. Bailev and E. H. Johnstox
(PLATE II)
0047 is an incomplete roll of the Stein collection at the India
Office, measuring 130 X 31 cm. The accompanying photograph
shows the roll except the lower part (32 cm.) containing six
short vertical lines (a business account) in Chinese, the last 10 cm.
being blank. The reverse of the roll contains the Chinese version of the
Aparimitajuih Sutra, and one line in Tibetan script. The obverse
here edited consists of Sanskrit verses with Khotanese Saka
annotations. There is a brief description of the roll by the late
Dr. A. F. R. Hoernle in Stein, Serindia, p. 1449.
The Sanskrit orthography ^ is similar to that of texts published by
Dr. Hoernle in JR AS., 1911, 447 ff. The Saka is of the late tj'pe found
in documents ..of the times of the kings Visa Dharma and Visa
Sambhava, about the end of the eighth century a.d. A terminus
ad quern is given by the closing of the Tun-huang library about the
beginning of the eleventh century.
The MS. is carelessly written in cursive Brahmi script, and the scribe
has struck through many of the aksaras. — H. W. B.
The Sanskrit text, unearthed by Dr. Bailey, contains a passage
from which important deductions may be drawn on a vexed question
of the history of Buddhist dogma. It falls into two parts, the first of
which consists of the opening verses of several works. Two of these,
the Mahdyanasutralamkdra and the Mulamadhijamakakdrihds, are
too well known to need comment, though the copyist distinguishes
himself by transposing the authors' names. Of the remainder, the
first is from an unnamed work, which I cannot identify but which
dealt perhaps with the “ false views ”, and the third is attributed to
the Mahdijdnasamdsa , a title apparently unknown to the Tibetan
and Chinese translations ; the application of the epithet nirnmla to
dharma suggests the possibility that it is a work of the Prajnaparamita
school. The last verse in this part is described as opening the
Ratnayotravibhdgasdstra of the Bodhisattva Maitreya, that is. the
’ It is not excluded that in the Sanskrit virama may have been at times intended
by the sign transliterated i, even with tt.
78
H. W. BAILEY AXD E. H. JdHXSTOX -
Uttaratantra, the crabbed Tibetan version of which has recently been
rendered into English with remarkable skill by Dr. Obermiller in
Acta Orientalia. ix. The Chinese tran.slation (Taisho Issaikyo ed..
Xo. 1611) is usually styled the Malidycinottaratantramslra. despite the
fact that the title literally translated, as pointed out long ago by
Xanjio, is Uttamilai/dnaralnagotrasdsfra. where ckaydna should
presumably be taken as the translator’s interpretation of the
.significance of the term tatitra. According to P. C. Bagchi, Le Canon
houddhique en Chine, p. 249, a Chinese catalogue of a.d. 597 knows
an alternative title, of which the first part is Rtitnagotravibhdga, in
agreement with the roll. The second part of the text is an excerpt
of nine verses from the same work, chapter iii. 1-7 and 9, according
to Dr. Obermiller. who has amalgamated the two verses, 5 and 6,
into one ; the copyist has also numbered the verses, but wrongly,
treating the Sardfilavikridita v'erse. number 7, as two, by reason of the
transcription dividing each pclda into two parts at the caesura.
To appreciate the value of the find. I must discuss the present
state of our information about the Uttaratantra. The Tibetan version
presents a work in five chapters and names Asahga as the author,
so that it is ordinarily reckoned among his five main works, which
were revealed to him bv the future Buddha. Maitreya. A curious
feature of it is the way in which kdrikua of the usual mnemonic curtness
are interspersed with verses or series of verses composed on regular
kdvya principles. The roll contains two good examples, in the
yathasainkhya compound of its verse 8. a figure several times employed
by Asvaghosa but not in favour with the poets of the classical period,
and in the repetition, amounting to internal rhvme, of Aha in the
last ixlda of its verse 10 ; and there appears to l)e a long series of such
verses in chapter iv. The Chinese tran.slation. which was prepared
liy Batnamati, a monk from Central India, early in the sixth century
A.U., gives no author's name ; but according to Peri, BEFEO., xi,
pp. 348-55, a seventh-century writer named Fa Tsang attributed
it to a certain Chien I and transliterated the name as Saramati, though
the more natural rendering of the Chinese, translation would be
Sthiramati. This autliority place.s Saramati in the 700 s and
Vasubandhu in the OOO's. while Paramartha. who accepts the same
period for Va.subandhu. places the other in tlu' beginning of the
goo's shortly after Xagarjuna. It is also to be noted that Saramati is
associated doctrinally with the supposititious Asvagho.sa. the author
of the so-called MahdydnasraddkotpddasdAra. Fa Tsang attributes
A FRAGIIEXT OF THE FTTARATAXTRA IX SAXSKRIT
79
to him also Xos. 1626 and 1627 of the Taisho Issaikyo Tripitaka. the
title of which is literally Dharmadhdtvavisesatasastra, and he is also
said to be the author of the Ju ta cFeng (usually translated
Mahdiidndvatdra) ; according to Peri this last work contains a single
allusion to the Mahdijdnnsfifralamldra, besides numerous references
to earlier works, hut. as it was translated before a.d. 440, the
identification of the (juotation may be incorrect or the passage an inter-
polation. For. if the work is really by the same writer, he lived,
according to the two Chinese authorities cited above, long before
Asahga. Eatnamati s translation of the Uttaratantra is a curious piece
of work ; though nominally divided into four chapters, it falls primarilv
into two sections. The first, covering pp. 813-820 of the Tai.sho
Issaikyo edition, consists solely of verses with occasional captions ;
the first thirty-nine columns are an ititroduction. and it then proceeds
to give a selection of verses from the text. Whether it has verses not
found in the Tibetan translation, I am unable to say. The second
section, beginning near the end of chapter i and containing in its
fourth chapter the end of the Tibetan's chapter i and the whole of its
remaining four chapters, gives the work with the commentary. The
latter differs in minor details from the Tibetan, hut resembles it too
closely not to have been translated (or perhaps better, paraphrased)
from the same original. This second part is remarkable in that it seems
to omit the kdvga verses of the Tibetan ; thus verses 8 and 10 of the
roll and iv, 14-86 of the Tibetan are not to be found in it. On the
other hand, verses 8 and 10 of the roll and some of the missing verses
of the Tibetan's chapter iv are given in the first section. The
explanation of this state of affairs may be that two separate tran.slations
have been joined together by the Chinese : for Bagchi. op. cit.. p. 250,
quotes a Chinese catalogue of the eighth centurv as .saving that
it is the second translation, and he remarks that Bodhiruci. who
collaborated with Eatnamati in another tran.slation. also translated
the Uttaratantra, though he does not identify it with anv of the works
of Bodhiruci mentioned in the catalogues, ihii/., pp. 252-260. The verses
common to both sections, so far as I have tested them, are worded in
identical terms, and one of the two .sections may well be by Bodhiruci.
The divergence between the Chinese and Tibetan traditions is
therefore considerable, and the value of the roll, which supports the
Chinese in its title and the Tibetan in its ascription, lies in the bearing
of its evidence on the puzzle. In the first place. Tibetan tradition,
as cited by Dr. Obermiller. recognizes the peculiar position which
80
H. W. BAILEY AXD E. H. JOHNSTON-
the Uttaratantra takes in the corpus of Asahga's works by virtue of its
strong tendency to Madhyamika tenets. On one point, however,
according to Dr. Obermiller's translation, it definitely accepts
Yijnanavadin doctrines, namely with regard to the Buddha-bodies ;
for, following the interpretation of modern Tibetan scholars, the
English version makes it teach the existence of three such bodies.
Now, besides the quotations from a number of texts in La Yallee
Poussin, Vijiiaptimatratdsiddhi, pp. 762 ff., the standpoint adopted
by the various sects and schools with regard to this doctrine is set
out in complete and masterly fashion in Hobogirin, s.v. busshin,
from which it appears that the Madhyamikas accept the existence of
two bodies only, and the Yijnanavadins, with the sole exception of
the treatise under discussion, three or more bodies. The Sanskrit
text, however, of verses 1-4 of the roll, which agrees in this point
entirely with the Tibetan text when properly understood, shows that
the modern Tibetan method of interpretation followed by
Dr. Obermiller is wrong as reading into the text a meaning it cannot
bear, and that the Uttaratantra in reality knows of only two Buddha-
bodies, the paramdrthakdya and the saihrrtilcaya. The author of the
article quoted in Hobogirin understands the Chinese version similarly
to teach the doctrine of two bodies only. With the disappearance of
the trikdya doctrine from it, there is nothing left to attach it with
certainty to the Yijnanavada. Two points in the Chinese further
support the view that we are dealing with a purely Madhyamika work,
firstly that Eatnamati and Bodhiruci's activities appear from the list
of their translations to have been confined to the sutras used by the
Madhyamikas and to the treatises based on them, and secondly that
Eatnamati, by translating tantra by ekaydna. indicates his view that he
is translating a Madhyamika work ; for, as Dr. Obermiller has pointed
out, the ekaydna theory of the gotra is a specific teaching of that school
as opposed to the Yijnanavadins.
Are we, then, justified in attributing a purely Madhyamika work
to Asahga on the strength of Tibetan tradition and of the attribution
to Maitreya in this roll, and in direct opposition to earlier Chinese
evidence ? In this connection it must be remembered that the latter
goes back to the school of Hiuan Tsang ; according to Demieville,
BEFEO., xxiv, pp. 52-3, Yuan Ts‘e, a pupil of his, ascribes the Pao
hsing lun (Ratnagotrasdstra) to Saramati, and, ihid., p. 55, n. 4, the Fa
Tsang, already quoted from Peri, certainly had some connection with
the same school. The famous pilgrim was deeply versed in the
A FRAGMENT OF THE UTTARATANTRA IN SANSKRIT 81
Vijnanavada, and it seems impossible that, if he thought the
Uttaratantra to be by Asahga, his followers should attribute it to a
writer of a different school. On the other hand, Dr. Obermiller supports
the Tibetan ascription by pointing out the identity of Uttaratantra,
i, 152, with Abhisamaydlanikara, v, 21, and the similarity of two verses
with passages in the Mahaydnas-fit rdla tid'd ra . The first of these cases
does not hold good. The verse in question, first elaborated by
Asvaghosa, was adopted and transformed by Xagarjuna, who read a
new meaning into it ; in that form it was quoted frequently in
Madhyamika and Vijhanavadin works (La Vallee Poussin, MUanges
cJiinois et houddhiqxes, i, p. 394). Moreover, I have been unable to
discover the verse in either section of the Chinese translation, and it
may be an interpolation. As regards the testimony of the roll, the
term ratnagotra is known to Asahga but used differentl}', Mahdydna-
siitrdlanddra, iii, 10 (we may compare the ratnakida of the Tantra in
Tibet, G. Tucci, Indo-Tibetica. III. pt. i, p. 43), and the expre.ssion
gumprabheda of the roll might be compared with gotraprabhedatd
of the same work, iii, 1. The style of the verses of kdvya tj’pe in the
roll suggests an earlier date than Asahga to me, though such a point
is too subjective to be relied on. Altogether the evidence in this
respect does not appear to me to have cogent force or to outweigh
Chinese statements, which on the face of them are well authenticated
and which give a result in accordance with what we should prhna
fade expect.
The final question is, if Tibetan tradition and the ascription in the
roll are wrong, how did they come to go wrong ? Now the restoration
of the name Saramati from the Chinese seems to me doubtful. It is
a somewhat unusual form and it does not really coincide with the
Chinese translation, Chien I ; further, Chinese transcriptions of proper
names have often come down to us in a corrupt state, and in the
present case the loss of a single character in copying would have been
enough to change Sthiramati into Saramati. The former, in my opinion,
is the correct form of the name ; for we do know of an early Sthiramati
from other sources, the author of a commentary on the Kdsya-
paparivarta, a very early Mahayana sutra, which is far older than the
Vijnanavada school. This commentary, No. 1523 in the Taisho
Issaikyo Tripitaka, was translated by Bodhiruci who gives no author's
name, but Baron von Stael-Holstein, in his preface to the Kdsya-
paparivarta, pp. xiv-x'vii. points out that the Tibetan gives his name.
In note 1 to the preface of his edition of the commentary he quotes
6
VOL. VIII. PABT I.
82
H. W. BAILEY AND E. H. JOHNSTON
also Bu-Ston’s History of Buddhism as authority for the statement that
this Sthiramati was born in the Dandakaranya. The reliability of this
statement is possibly open to doubt, but it is curious that Eatnamati,
the translator of the Uttaratantra, also came from Central India.
The theory I would put forward for consideration is therefore as
follows : There was a writer named Sthiramati, who lived some time
before Asanga, and who wrote the Uttaratantra or Ratnagotravibhdga,
the commentary on the Kdsyapaparivarta, the Dharmadhatvavisesatd-
sdstra, and possibly the Mahdydndvatdra ; he belonged to the
Madhyamika school, but to a section teaching a doctrine more advanced
than that of Nagarjuna, of a type approximating to that of the
mysterious Mahdydnasraddkotpdda, and in some degree anticipating
Yijnanavadin developments. As he failed to found a school which
could compete with the two great divisions of the Mahayana, his works
fell into comparative oblmon at a fairly early date. Later, after
Asanga and either after or contemporaneous with the date of the
Chinese translations of the first Sthiramati's works, there was a much
more famous Sthiramati of Valabhl, an adherent of the Vijnanavada,
who YTote a number of commentaries, some stiU extant in Sanskrit.
By the time of the roll, somewhere according to Dr. Bailey in the eighth
to tenth centuries a.d., when the name Sthiramati was found attached
to the Uttaratantra, it was supposed to be that of the only Sthiramati
whose name was still Hving, namely Asanga’ s follower and com-
mentator. It would be natural in such circumstances to suppose the
commentary alone to be by him, in which case the original, whose
teaching shows some affinities with Vijnanavadin doctrines, could only
be attributed to Asanga. The word tantra in the title might assist in
the mistake, as Asanga is traditionally associated with the origin of
that school. This theory has the advantage of accounting for all the
facts and of making the position of the Uttaratantra in the history of
Buddhist dogmatics much more comprehensible, and the same cannot
be said of any other possible explanation. Further than this the
matter can hardly be taken, till more evidence is produced, for instance
by a much completer examination of the Chinese translation than I
am competent to undertake.
A few words must be added about the restoration of the Sanskrit
from the transliteration. The original, which cannot always be read
with certainty, has a certain number of corruptions, besides a number
of extra syllables, which should have been cut out and which I have
omitted from the Sanskrit ; the method of transliteration also in the
A FRAGMENT OF THE UTTARATANTEA IN SANSKRIT
83
roll is not uniform, particularly with regard to the vowels, while the
language is such as to make conjecture unsafe. As it is not possible
therefore to give the Sanskrit that was intended by the writer by relying
solely on the MS., I have made use for the Uttaratantra of the Tibetan
translation in Mdo 45 (India Office copy), and of the Chinese translation
(verse i, 1, on p. 820 c ; \w. 1-6 and 9 on pp. 844 a and h ; and vv. 1,
4, 6, 8. and 10, on pp. 817 a and 6), and this has enabled me to prepare
a text which, subject to one or two small matters of uncertainty, I feel
justified in considering to be that which lay before the author of the
roll. The Tibetan has occasional variant readings, which I give under
the letter T in the notes, but there is in reality almost complete
textual agreement between the three versions, the Chinese being far
more literal than is often the case.^ — E. H. J.
^ Since the above paper was written the Kev. Kahiii Saiiikrityayana has announced
in xxi (193o), pp. 31 and 33, the discovery in Tiln't of two incomplete MSS.
of the Sanskrit text of the I'ttnmUintra. It is to be hoped that photographs will be
made available for the publication of this important material.
84
H. W. BAILEY AXD E. H. JOHXSTOX —
Ch. 0047.
1 . [Fragment of loicer part of alsara da.]
2. ragadidaursam - visadrrigdi - kudrristisalyam =
ganidabhineddham - hrra = dayam samhTriksi
3. laikam = ttrattu paramsrrimimagatta karinattmimme%a
yam = stasidkya^muhta = ma^ - bhaisaguri
4. mabhyarca^yanime = arthajnarthivlbhavina
prriloiratte va va ® va “ padistattima
5. lai = dukhittisyattittirinayi dukhittaujine =
karumnastattinta^tti^mayain
6. sutralamkarisastri lilya padauysa karya - arya
nagarrjum baudhisatvina hva
7. sarvajnayi namahi : stasiiiu : parartha t ’
bliaritta'ttiinane = uarimala
8. yi ci dhannayani = dimattiginivarayi ca =
mihayainiia siniasisa*stri® hlya padau-
9. ysa karya =
anaraudha anautipada • anachidaknanasitta =
manekartha + mana
10. karthamanargimanirgima = yatti
prrittittisimuttpada prriparncapamcima
11. siva = desa®me^yam ® niahasambudhainstaravaiu
ku de^va^tta ^ baudhattauvara” = liaiidhatta- -
12. madhyimai sastri blya padauysa karya arva mittrai
baudhasatti jsa livata
13. sti - dharmasta - ganasta diuittii = baudliariguni
karmarn cd baudhainevarn = krrisijasva
^ samrrifta written over iidahhtnf .
- Struck out.
^ With ~i stnitk out.
* muhfa=ma below the line.
* With -« struck out.
Below the line.
" Uncertain, possibly ra,
® With -au stniek out.
A FRAGMENT OF THE FTTARATANTRA IN SANSKRIT
85
Sanskrit Text, with English translation op Saka
(1. 2) Eagadidosa\Hsadigd]iakudrstisalya- ^
gandabhinadcDialirdayaih samaveksya (1. 3) lokam |
Trata ^ parasrayagatan karunatmako yas
taiii siddhimuktam avisam ^ guru(l. 4)ra arcayami 1|
Arthajno ’rtha\dbhavanam prakurute vaca padais cama(l. 5)Iair
dubkhasyottaranaya * duhkhitajane karunyatas tanmayah |
The first karika of the Sutralarnkarasastra. Spoken by Arya-
Nagarjuna.
(1. 7) Sarvajnaya nanias tasniai panirthaprahitatmane |
Nirmala(l. 8)ya ca dharmaya dhimadganavaraya ca |1
The first karika of the ilahayana-samasa-sastra.
(1. 9) Anirodhani anutpadam anucchedam asasvatam ® |
Anekartham ana(l. 10)nartham anagamam * anirganiam H
Yah pratityasamutpadaih prapahcopasamarh (1. 11) sivam |
De^ayamasa saihbuddhas ’ taiii vande vadatahi varam H
The first karika of the Madhyaniaka-sastra. Spoken by the
Bodhisattva Arya-Maitreya.
< Buddha > (1. 13)s ca dharnias ca ganas ca dhatur
bodhir gunah karma ca bauddham eva ® )
^ °salyarh would be preferable metrically.
- Or tratva.
® The restoration is doubtful.
^ The roll reads wrongly duhkhitasyo^.
^ Did the writer intend aw«5<Vam ?
® It looks as if an<igrimam is meant here.
^ Is desayan inahdsarhbuddhas intended ?
® tha-viQy T., i.e. ante.
86
H. W. BAILEY AND E. H. JOHNSTON
14. sastrasyi siriraniettatti = simasatta
vajfia padhamna saptta =
mittrai baudhasa -
15. t^d jsa hva - rahna = gautmibhagisastri hlya
padauysa karya
16. . . svartba-parartha parimarthika yam =
stidasritta samvrrittakajnntam ca - ba ®
17. lam \'isamyaugani \dpakabhavanitti^®
ettatti citturisastigunam prribhedam = 1
18. svarthasampattA’idrristamta = mrrise
saketta sirlram parimarthika = parisara
19. pattidrristairmta = mrrise sakettakam \dpu 2
\isamyaugagunlr}dimii = ttam = \dpu
20. radvdbaladabai = vipakakam dvlttlyaraca^^ =
mahattipurusalaksarnnam 3
21. balarn tvavajnarnna-VTrittesa vajnavatti =
stidasrritta sam%Trittakayattam ca
ttithagattavirnpakattantii^riksivatti^® =
22. visaradarntva parasatti.su slha : vatti = ukatta
23. muninldrrida darsanamambhu^®camdrrivatti = 4
sthamnasthaniiia vipaki ca = karma
24. namirndrriye*^sa ca = dhatvesamldhamuhttau ca =
miuiiirge sarva*®ttrigrammane 5
25. dhvaninadakidesa\’imalya = nivasranu
prasiinrrittei^ysamntha smrrattevl+he
26. = devicaksusamttarn ca = jnannaTn
disividarn balam = 6 stharnnasthanina vl
® bd with a probably struck out.
Struck out.
ryu written under ru (with w struck out).
tta struck out, ca written below.
sti . . . ca struck out, = I. 16.
ri with ’i written above,
tti . . . vatti below the line.
hu struck out, bhd beneath.
An uncertain ak’^ara struck out, ye written below.
rvi with -i struck out.
prasimrritte struck out,
y.^am ntha uncertain.
ttrdm struck out, with ttdm beIo\v.
A FRAGMENT OF THE UTTARATANTRA IN SANSKRIT
87
Krtsnasya (1. 14) sastrasya sarlram etat
samasato vajrapadani sapta Ij
Spoken by the Bodhisattva Arya-Maitreya. The first karika
of the Ratnagotravibhaga sastra.
1. (I. 16) Svarthain parartham paramarthakayas
tadasrita sariivrtikayata ca |
Pha(l. 17)lairi visamyogavipakabhavad
etac catuhsastigunaprabhedam [j
2. (1. 18) Svarthasampattidrstantam ® sariram paramarthikam |
Parasarii(l. 19)pattidrstantam rseh samketikarh vapuh 1|
3. Visamyogagunair j-uktam vapii(l. 20)r adyam baladibhih |
Vaipakikarh d\dtiyam ca mahapurusalaksanam I|
4. (1. 21) Balarh tv a\djnana%’Ttesu vajravat
Tathagatavenikatantariksavat |
(1. 22) Visaradatvam parisatsu s'rhhavan
(1. 23) muner dvidhadarsanam ambucandravat ||
5. Sthanasthane \'ipake ca karma(l. 24)nam indriyesu ca |
Dhatusv evadh'muktau ca marge sarvatragamini |1
6. (1. 25) Dhyanadiklesavaimalye nivasanusmrtav iha |
(1. 26) Divy'e caksusi santau ca jnanam dasavidharh balam |1
• T has bdag-nid-kyis hbyor-ba, possibly for siatasampatti° .
tm-rigs sgnbs-pa, T, i.e. avidydrara^esii, which is the better reading.
The Chinese and T transpose b and c, rightly as the continuation shows. The
three following characters do not belong to the verse and may represent a misplaced
uktam.
T has Idan-pa “ possession ” ; the only possible word, prdpti, is metrically
difficult. For paltau from pad, which is free from this objection, there is only the
authority of the Indian lexicons.
88
H. \V. BAILEY AND E. H. JOHNSTON
27. pakadattusi iagatti= nanadhimuhttimdrriyetti =
cikrra^^vari^®rtti
28. manayaugabkumesa jagatti = pur^Venavaprra^®smrrattau
karumnastattittatti^^mayam ga-devye caksau
29. casavakslyikrrittau = vajSamnavarmacalain =
prrakaramdhummabhedaprrika
30. ranam = chaidbaautti balavajrravatti = 8
sarvadharma\nsainbaudham = vibamdau
31. prritta sedhana = margarya dani raudhattpl = pti
visaradhjd citturividdbam = 9
32. naittya va^®nauntesa yitha mrrigrrimdrra =
ne ra^^rbblranumtha grritte mrri
33. gebhya= munemdrrasihaupi statha
gane^^sa = svasthau narastha sthira
34. vekeramastha = 10
^ysumnara tcauraksastyam pajam dharmam u
tcihauryam visa -
35. radhyam-rahna^- = gauttravibhagisastri jsi
niramda
36.
37. punam saram rya gam ?
38. =puna samrraraurya ga ?
Six vertical lines of Chine.se follow.
-- krri with A struck out,
23 ri struck out.
2* j>u . . . ttau below the line.
Uncertain.
2® ttatti below the line ; karitm . , . mnyam struck out.
2’ rya first written, later stroke radded to left.
2® ptl below the line in thinner hamlwritintr.
2® vi with -i struck out.
2® ra struck out.
Uncertain aksara struck out, with ne below.
22 rahna to end of line struck out,
22 ram with ryarri below.
2^ uncertain sign.
A FEAGMEA’T OF THE t'TTARATAXTEA IS SASSKKIT
89
8. Sthanasthanavi(l. 27)pakadhatusu jagannaiiadhimuktliidriye
cakravarta(l. 28)nayogabhumisu jagatpurvaparanusmrtau |
Divye caksusi (1. 29) casravaksayakrtav ajfianavarmacala-
prakaraiidukabhedanapraka(l. 30)ranaccbedad balam
[vajravat ||
9. Sarvadharmabhisambodbam vibandha(l. 31)pratisedhanam |
Margajnata nirodho ’pi vaisaradyam catur\'idliani
10. (1. 32) Nityam vanantesu yatha mrgendro
nirbbir anuttrasagato mr(l. 33)gebbyah ]
Munlndrasimho ’pi tatha gariesu
svastho nirasthah st!iira(l. 34)\'ikraniasthah 1|
Of the sixty-four diidne special qualities and the four intrepidities.
Excerpted from the Ratnagotravibhagasastra.
{Translation uncertain.)
non-moiis rnam-hyaix, T, i.e. samkU-iarynvadana', It omits bhumisn and jagat,
and reads °puriapard'. The Chinese had hhiimt and paraphrases the preceding words
“ all kinds of tendencies {suif Giles 10396, ? anui>aya) and abandonments {prnkdna) ”,
reading perhaps citrd^ for cakrd^.
prakarana from kJ\ certified by the Chinese's ^'^av ‘‘disperse" ; one would expect
pradaranOf possibly indicated by T’s gzhig.
legs-gna-s, T. i.e. snstko ; so also the Chinese.
Two Vedantic Hymns from the Siddhantamuktavali
By Axanda K. Coomaraswamy
^pHE {V eddnta-)SiddhdntamukldraVi of Prakasananda. a work of the
sixteenth century, written in refutation of Ramanuja’s Tattvasdra,
has been edited and translated in another manner by A. Yenis in the
Pandit, N.S., vols. 11 and 12, and reprinted separately, Benares, 1912.
In our version we have endeavoured to be absolutely literal without
employing the customary phraseology of Sanskrit scholarship, some
of which is actually misleading, and other parts obscure to the general
reader. On the other hand, we have not he.sitated to employ the
technical terms of scholastic philosophy in their proper context ;
we maintain, indeed, that the content of Indian religious or
philosophical texts cannot be conveyed in any other way ; and that
the propriety of this procedure will be apparent to anyone precisely
to the extent that he is familiar with both Hindu and Christian
scholastic method. It is not intended that the result make for easy
reading ; on the contrary, the modern reader, accustomed to the use
of words in vague or much degraded senses, and to the making of hasty
assimilations, must be faced with the necessity of establishing for
himself the content of unfamiliar references, which is even more
essential here than it would be in the analogous case of the study of the
Latin hymns of the Middle Ages. It will therefore be understood that
the translation is a technical one, and to be taken accordingly. We
beheve at the same time that by these means the formal beauty and
clarity of the original are better preserved than would be possible
in an easier, vaguer, and more sentimental wording.
What may be found remarkable in the translated hjnnns is their
markedly devotional character, which might not have been anticipated
in connection with so highly technical and controversial a treatise
as is this “ Clew to Freedom, the Last End and Whole Intention of
Revelation ” ; it is not really more remarkable than that St. Thomas
should have written not only the Smnnm Theologica and Summa
Contra Gentiles, but also the hymn Adoro te devote. In “ man’s last
end ” the Way of Gnosis (jnana yoga) and the Way of Participation
(bhaliti yoga) meet in the full consent of Intellect and Will, where Being,
Intelligence, and Beatitude — the equivalents of the more familiar
92
AXAXDA K. COOMARASWAMY —
“ Goodness, Truth, and Beauty " — are no longer thinkable as
accidental attributes (iijxilaksann) annexed to the First, but only as
essential attributes (visesam) or names of the First, idem in seipso.
I
Ya atma sarvavastunam yadartham sakalam jagat
anandabdhih svatantro ’savanadeyah katharn vada.
yadanyad vastu tatsarvarn yadbhede narasrngavat
sata sarvapadarthanani anadevah katharn vada.
yadvase praninah sarve Brahmadyah krmayastatha
Isanah sarvavastunam anadeyah katharn bhavet.
yaccaksuh sarvabhutananr manaso yanmano viduh
yajjyotir jyotisain devo nopadeyah katharn vibhuh.
modapramoda-paksabhyam anandatma tamo gatah
jivayatyakhilan lokan anadeyah svayarn kutah.
yasyananda-samudrasya lesamatram jagadgatam
prasrtain Brahmalokadau sukhabdhiin kah parityajet.
hairanyagarbham aisvaryain yasmin drste trnayate
sIma sarvapumarthanam apumarthah katharn bhavet.
yatkama brahmacaryanta Indradyah prapta-sampadah
svasvabhogarn tyajantveva na pumarthah katharn nrnam.
yaddidrksaphalah sarva vaidikyo vividhah kriyah
yagadya vihitas tasminnupeksa vada te katharn.
yaddrstimatratah sarvah kamadya duhkhabhumayah
vinasyanti ksanenasavupadeyah katharn na te.
ahladarupata yasya susupte sarvasaksikl
tatropeksa bhaved yasya tadanyah syatpasuh katharn.
II
Anandarupam atmanain .saccid advaya-tattvakam
apurvadi-pramaiioktarn prapyaharn tadvapuh-sthitah.
Yo ’ham advaya-vastveva sadvaye drdha-niscayah
prapya canandam atmanarn so ham advaya-vigrahah.
nasti Brahnra-sadanandam iti me durmatih sthita
kva gata sa na janami yadahani tadvapuh-sthitah.
purnanandadvaye tattve Mervadi-jagadakrtih
bodhe 'Irodhakrtaivasld abodhah kv'a gato ‘dhuna.
sainsara-roga-sartigrasto duhkha-rasir ivaparah
atma-bodha-samunmesad anandabdhir aharn sthitah.
vo ’ham alpepi visaye ragavan ativihvalah
anandatmani samprapte saragah kva gato ’dhuna.
TWO VEDANTIC HYMKS
93
yasya me jagatam kartuh karyair apahrtatmanak
avirbhuta-parananda atma praptak sruter balat.
paramrsto ’si labdho ’si presito ’si ciram maya
idanim tvam aham prapto na tyajami kadacana.
tvam vina nihsvarupo ’ham mam vina tvani katham sthitah
distyedanim maya labdho yo 'si so ’si namo 'stu te.
dehabhimana-nigadair baddho ’bodhakhya-taskaraih
ciram te darsaiiad eva trutitam bandhanam ksanat.
visuddho 'smi vimukto 'smi purnat purnatam akrtih
asamsprsya mamatmanam antar Brahmandakotayah.
tattvam-adi-vaco-jalain avrttam asakrt pura
idanim tat sravad eva purnanando vyavasthitah.
I
How can it be asserted that that Essence tliat is the essence of
every substance, and iii which the whole world subsists as in
its sufficient reason, that authentic Ocean of Beatitude, is other than
man's last end !
How can that Essence that is the ground of all else, which if
divided from that whole is as non-existent ' as the horns of a man, how
can that Essence which is the reality of all categories, be other than
man's last end !
How can That, within whose sway are all things living, from Maker ^
to emmet, and being the ground of all things is their Ruler, be other
than man's last end ?
How can That, which Comprehensors call the Eye of all things, the
Intellect of intellects, the Light of lights, and numinous Omnipresence,
be other than man's last end ?
How can that Essence and Beatitude which, when it enters the
Dark-world on its wings of enjo^unent and satisfaction.^ quickens everv
world, be made out to be other than man's last end ?
MTio can refuse that Ocean of Beatitude, of which but a little
measure pervades the world. ^ and is yet the Tide of Well-being that
flows in the 3Iaker's heaven and every other world ?
How can That, which when it is seen, the lordship of the Golden-
Germ is valued at a straw, and which is the uttermost limit of human
value, be other than man's last end ?
How can That, for which Indra and others, though each was
complete in his own manner, abandoned their several modes of enjoy-
ment, becoming anchorites, be other than man's last end 1
94
ANAXDA K. COOMARASWAMY —
WTiicli to yearn to behold is the fruition of all ritual acts and
appointed sacrifices, how may it be said that therein is aught but
man's last end 1
How can That, at sight whereof all longings ® whatever, the bases
of ill, are suddenly destroyed, be other than man's last end 1
How is he aught but a brute beast who refrains from that Form of
Refreshment that is immediately and universally seen in Deep
Sleep 1 ®
II
Xow that by means of the norm ’ that is now as it ever was I have
found the Essence that is the Form of Beatitude, and other-less
Principle of Being and Intellect, I too subsist in that likeness.
That same I that was firmly persuaded of otherness, albeit really
devoid of otherness, now subsist in the form of Xo-otherness.®
This was my silly thought, to wit, that “ That Omnipotence, Being,
and Beatitude is not " ; but now subsisting in that very image, I
know not where that fancy fled.
The outward seeming of the world of Meru and so forth was wrought
unawares in the very principle * of plenary Beatitude ; but now that
I wake, where is that slumber gone ?
Clutched by the fever of the flux of things,^® I was estranged, and
naught but a mountain of ill ; but now by uprush of awareness of the
Essence, I subsist as the Ocean of that Beatitude.
That Essence erst was rent away by the working of what in me was
the maker of semblant worlds ; but since by power of the self-revealed-
Word I have found the Essence, Beatitude Supernal is made apparent.
Thou hast been touched and taken ; long hast Thou dwelt apart
from me, but now that 1 have found Thee, I shall never let Thee go !
Without Thee, I have no being in myself ; without me. how mayst
Thou be ? Hail unto Tbce ! Thou that art That that I found by
transmitted doctrine !
Bounden I was in shackles of imagined flesh by bandits of delusion ;
but now at last by mere sight of Thee the bondage has been broken
suddenly.
I am cleansed and enlarged, in a likeness more plenary than any
plenum ; and what though mjTiad worlds inhere, there is none can
touch my Essence.
Time was, “ That thou " and such like sapngs were hidden as
though by a net, and so it ever went ; now plenary Beatitude subsists
in audition of that very Word.
TWO VEDANTIC HYMNS
95
Hymns of this sort could be annotated at indefinite length, but
here we shall notice only expressions the values of which can be
elucidated by a brief citation of analogous texts : —
1 “ Non-existent ” : things considered apart and as they are in themselves
(ordo per esae) are naught when compared to the Essence in and whereby tht_\- subsist,
cf. St. Augustine, Confession um, xi, 4, ijno compatalo ner. pulchra sitnl, nt’c hona siiiil,
nec sunt, and ib. esse quidem, quoniam abs te suiil, non esse autcw, quoniain id quod
es non sunt, i.e. “ Compared with Whom, things are neither beautiful, nor good, nor
are they at all ”, and “ A being they have, because tliej' are from Thee : and yet no
being, because what Thou art, they are not.”
Yad-bhede narasrnqarat also corresponds to St. Thomas, Quaest. disp. de veritate,
q. 8, a. 7, ad 2, Omnis creatura . . . si consideratur sine hoc qvod ah alio habet, est nihil
et tenebra etfalsitas, cf. Eckhart, Super Oratione Dominica, temporalia, maxime respectu
aeternorum, nichil swit. St. Thomas, indeed, is careful to add that non est iutelliqendum,
quod essentia sva sit tenebra vel falsitas, nor does thi.s differ m an}' respect from the
Vedantio position, which likewise asserts the absolute reality of the dtnian which is
the vastu of all things.
The expression " man’s la.st end ” is taken from the text immediately preceding
the hymn, which i.s an answer to the false assertion that “ the Essence is not to be
taken as man's last end ” (atniauo' piiiusdrlhatie prdpte), the opponent maintaining
that “ man’s last end is merely the cessation of ill ” {duhkhdhhdvdrtham eia). The
actual refrains give only latham lada “ How can you assert ? ” (or variants of these
words), a following dtmano' puriisdrthatiam being understood.
The rendering of dtman by “ Essence ” and “ essence ” (respectively paramdtman
or antardtwan, and pratyagdtman or adhijdtman), which I adopt henceforth in place of
the customary " Self ” and “ self ” (awkward in English in various ways, and especially
because of the connotation “ selfish "), has great advantages, both as being more
exact in reference and as facilitating comparisons (the problem of the identity or
distinction of the divine and human essences being treated at length in Christian
theology, which asserts their distinction, and in this respect diverges from metaphysics
according to which tat team asi). Essence (essentia) is that by uhich a thing is (habet
esse) in any mode whatever. Adopting the translation ' essence ” we have then
a jierfectly logical trilogy of dtman as that by which a thing is ; ndina, or “ form ”,
as that by which a thing is what it is ; and liijm, the phenomenal aspect, in which the
thing is as it is. The same trilogy corresponds to the Christian “ body, soul, and
spirit ” — ” matter, form, and essence ", and again to the Hindu concepts of the
anna-maya, mano-maya, and dnanda-maya “sheaths", the mdmsa, dirya, andjiiana
” eyes ”, and like formula:. The identity of “ essence ” with ' lifc-brcath ” or
“ spirit ” will be evident ; dtman, like prdno, derives from an " to breathe ”, or perhaps
from id " to blow ”, and the consubstautiality of essence and spirit is constantlv
asserted or implied in the Vedas, e.g. Sg Veda, vii. S7, 2, dtmd te idto ” the Gale that is
Thine Essence ”, or wherever Matarisvan is referred to as the kindler of Agni, Who is
in fact “ self-kindled ” or, better, “ kindled by His own Essence ” ; cf. Sg Veda, x,
129, 2, where " in the lieginniiig. That One ” (the same as Mrtyu, not yet dtnianvi,
in Drhaddranyaka Up., i, 2, 1) “suspires without spiration naif ai'dta — aprdnnh ,
ilaitri Vp., vi, 2fi). Rendering dtman thus by " e.ssencc ”, we reserve “spirit”,
" breath “ gale ”, " life ", etc., for pi ana, idta. idyii. Assumine at man to be Essence
the Buddha's last words take on a new .significance, thus : “ Be ye such as have the
Essence for their light, the Esseme for their refuge, and no other.*’
The rendering of dtman by " es.sence " is referent ially correct, inasmuch as the
dtman is preci'ely that unchangeable reality which underlie.s all accidents. The
rendering is nevertheless experimental, and it may well prove better to render
96
ANANDA K. CUOMARASWAMY
atman by Spirit or Spirant, in closer accordance with the etymology, and at the same
time avoiding the difficulty that is occasioned by the employment of essence univocally
with respect to things as they are m themseivo ami things as they are in God. Atman
as Brahman is coincident ith a Ijeing that is not iii any mode ; ntynan in the individual
is not that individual’s essence ", but the qua non, of an hahere or existence
in any mode. In verse 1, then, we might have read '* that Spirant that i& the spirant
in every substance ”, or even " that Life that is the life in every substance bearing
always in mind that Essence, Spiration, and Life are I'l^m m >eipso, one and the same
in the subject referred to, although apparently manifold in the worlds of ■* knowledge-
of’’ {cv:idya).
2 We render Brahma (m.) by " Maker the Creator, God as Deity in actn
being intended ; and similarly in the sixtli verse.
3 Dark-world " renders : from the Vedic point of view the creation is
essentially a penetration of the Dark-world by Light, the Supernal Sun “ releasing
all things in their kind " (t i.h'Ct riiiiuni prati mvhcatey v, 91, 2), which were veiled by
Darkness (fama’^i, x, 129. *3) in the beginning (ib. and J^ladri Up., vi, 2).
The concept of proccsshm upon wings corresponds to that of the divine procession
as symbolized in the Bg Vnla, e.ir. iii, ”>4, 8, ra/at pnfatri “ He proceeds on wings ”,
and vi, 0, 5, “ Intellect is the swiftest of birds,” while A^ni’s motion is several
times described as a falcon or eagle's swoop (cf. blooinfield, “ The legend of Soma and
the Eagle,” JAOS. xvi, 1, pp. ll-Io). To represent the Spirit and all Angels as birds
is common to tradition universally; in Cbristianity, for example, the Holy Ghost is
represented by a dov(* (and this " dove ” is the same as that dove which when it leaves
tlie ark finds that the Tree of Life has appeared above the ebbing Waters, and ‘‘ does
not return ”, but pen.lies there, being, as iMntc expre.sso.s it, “ the power that i.s form
unto the nests,” J^fnadi^o. xviii, llo, an image eon.stantly recurring in the Rg Veda
and Upani'^ad's, e.g. Bg V<da, x, 01, 2, where Agni '* like a bird makes His home in
every tree rane \anp h\is)bje faLvir irn, and Maitrayan'yti Tp., iii, 2, “ He fetters
himself like a bird in the net
’’Enjoyment” and ’’satisfaction” are employed respectively to render moda
and pifUtKida ; the terms are placed in apposition to ’’wings” in the dual, and
it is clear that a distinction of meaning is intended. We take it that there is allusion
to the well-known figure of the two birds that perch in the Tree of Life, one eating of
its fruit, the other looking on {Rg Veduy i, 164, 20-21), and that moda and pramoda
represent their respective modes of enjoyment ; the figure of the two birds being
replaced by that of the two wings nf the single bird. Our view that pramoda refers to
the vision of the whole {usmyti nhhirasfp, Rg Vrdo. i, 164, 44, ete.b and moda to that of
the parts, is confirmed by the use of the exprc'.’sion pramvdam prayati in Sahkaraearya’s
Si'dtnianirupnna, 9o, where it is said that the E'^senee, regarding the world-picture
painteil by the Essen''-e on the canvas of tin* Essence, experiences a great satisfaction
A close parallel can be found in Genesis, where at tlie end of each day's work God sees
that it is ” good ”, but \\h('ri all is rlone, that it is “ very good ”, cf. St. Augustine,
Go ///("'-■ S' ' o // ■?, Mii, 2S, '‘Thou sa\\e.‘'t <*verything that Thou hadst made, and behold
it was not only Good, but also Very Good, as lieing now altogrthe*r.”
It is also of much .significaiuo that the divine procession in both aspects
(adhidairata and adhyatnia) is rcfcrrerl to as essentially and altogether a blissful
experience (cf. Rg Veda, vii, 87. 2, ” Like an untamed hart that takes his pleasure in
the pa.stures,” and Eckhart, ” The joy and .satisfaction of it are ineffable ”) ; where
wc might perhaps have expected the assertion of an cxjierience of pleasure and pain
as the twin wings of pro( cssion, and might have looked for a corre.spoiifling cnnternptus
mundi. There is evidently assumeil, on tlie contrary, an intrinsic perfection of all
things ; which jierfection if realized in so far as they are known essentially and not bv
their accidents ; it is a.s?,umed, in other words, that the Compreherisor’s {vidrnn)
knowledge of ill is like the divine understanding wholly .vdi .spec/e boni.
The notions of the two wings ’ and of moda and pramoda are evidently derived
TWO VEDAXTIC HYMNS
97
from Taithriya Vp., ii, 2, where the wings are identified with prana and apana in the
prdna-maya hypostasis, and li, 5, where rnoda and pratnoda are the right and left sides
of the anandn-mayn substance of the superessential Person. In Bg Veda, x, 14, 16,
the Only Great {ekam brhrtt) “ flies into the six worlds
‘ One of the many Vedantie texts in which the Transcendence as well as the
Immanence of the Essence is asserted. The SiddhdiitamuktdiaU itself elsewhere
(pp. 67-8), assuming Immanence (vyapalafia) and Universal Presence of the Whole
Essence (sarvasamhandhatvarh sandtmand), shows that Transcendence is necessarily
implied ; for if one finite thing coidd be pervaded by the whole essence of another
finite thing, the two things -nmuld be one and the same ; therefore that Essence which
is assumed to be wholly present to every finite thing must be itself an Infinite
Essence.
In the course of the argument it appears incidentally that finite things are by no
means regarded as wholly non-existent (amt), but only as false (mithya) in the sense of
being inexplicable (anirracaniya) if considered apart from the Essence that is their
support (adhisihdna). It would be obvious in any case that amt could only ajiplv to
“ things ” to the extent that they are not wholly in act, an entity {sattva) being by
hypothesis what is in act and therefore has being (sat). Cf. St. Thomas, Sum. Thex)}.,
i, q. 16, a. Ic, “ Everything is said to be true absolutely in so far as it is related to the
intellect from which it depends . . . natural things are said to be true in so far as they
express the likeness of the species that arc in the divine mind,” and ib. q. 17, a. Ir,
“ In relation to our intellect, natural things which are compared thereto accidentally
can be called false ; not simply, but relatively,” i.e. with respect to our mis-
apprehension of their true nature : “ a thing is said to he false that naturally begets
a false opinion,” for example, when we call “ tin, false gold ” ; which corresponds
exactly to the Vedantie exemptum of the snake and the rope ; and ” although the
falsity i.s not in the thing but in the intellect ” (which represents the c(iuivalent of the
^ cdantic theory of avidyd), St. Thomas would certainly have agreed to call natural
things ” false ” in so far as they are mistakenly regarded as independently self-
subsistent. Thus it appears that it is extremely difficult, if not impossible, to state
a distinction of the Christian from the Vedantie doctrine in the matter of external
reality.
We venture to add, without being able to cite authority, that to assert an absolute
or simple “ unreality ” of “ things ” would be not only a contradiction in terms
(“ real ” deriving from res, “ thing ”) but also false in fact, in that a denial of reality
to thing.s would be (as .suggested above) to assert their existence merely in potentia,
whereas by hypothesis “ things ” are alwaj's at least partly in act, or else are not things.
“ JIaker ” renders Brahma ; and coresponds to “ Golden Germ ” in the next
verse. Hiranyagarbha (the ” Golden Germ ”) = Agni-Prajapati, the Year, the Sun,
etc., i.e. God in arlii. or as correctly understood by Sankaracarva. snquna, rnarlya,
nparn Brahman, i.e. Brahma. Cf. Kathn Up., ii, II, where Xaeiketas surrenders the
heavenly world that is the natural term of the deraydna for the sake of the Supreme
Identity altogether without otherness. It is a well-known Aupanisada (and Buddhist)
thought that man's last end is not accomplished with the attainment of a Heaven,
in which all desires arc fulfilled, but lies beyond ; and it is repeatedly asserted that the
manifested, i.e. manifestable aspect of deity, represents but a ” little fraction ” of the
whole divine being, which can only be grasped, if at all, by the rm rnnntionis (neti,
neti, etc.), the possibilities of non-manife.station infinitely transcending those of
manifestation. This relative disparagement of deity in act is exactly paralleled in
Eckhart (see i, 274-8 in the Evans edition), when he s,ays “ the soul has got to die to
all the activity of the divine nature if she is to enter the divine essence where God is
altogether idle ” : here “ where God is altogether idle ” = sdtUa dtmani, “ in the
Essence at rest,” Kntha Up., iii, 1,3.
^ “ All longings,” kdmdh, cf. Brhadaranyaka Up., iv, 3, 21-9. Kdmdh here also
corresponds to early Buddhist tanhd, as the origin of ill.
VOL. vni. PAST 1.
7
98
ANANDA K. COOMARASWAMY —
® “ Form of Refreshment,” aMada-rupata. “ Deep Sleep ” {sitsupla) is a technical
term in the well-known classification of modes of consciousness as “ Waking ”,
“ Dream ”, and “ Deep Sleep ”, designating the worldly, angelic, and divine modes of
understanding; Deep Sleep being synonymous with samudhi, where the distinction
of knower and known no longer confuses the understanding, or, in other words, where
knowledge-as perfects the imperfection of knowledge-of. Dharana, dhydna, and
samadhi (= sumipta) in Yoga correspond to St. Bernard’s consideratio, contemplatio,
and excesses or raptvs, but with this distinction, that the Christian excesms is
usually an ec-stasis, the Hindu somddAi rather an “ in-stasis
Note the correspondence of suaupla with susupdna characterizing Ahi-Vrtra in
RV. iv, 19, 3 ; the significance is developed in my ” Angel and Titan ”, to appear at
once in the JAOS. ; cf. also note 9.
“ Brute beast,” pasu, i.e. less than man and only fit to be offered up in sacrifice
by others ; cf. Brhaddranyaka Vp., i, 4, 10, “ Whoever worships any Angel as other
than the Essence, thinking ‘ He is one, and I another ’, does not understand, and is
like a beast for the Angels,” and Aitareya Aranyala, ii, 3, 2, where pasu is defined
as “ whose discrimination is merely by hunger and thirst ” (asartn-pipdse era
abhivijndnam), i.e. whose understanding is merely empirical and estimative. In
Aitareya Aranyaka, loc. cit., the distinction made is of purusa (person) from pasu
(animal), cf. Boethius, Contra Evtychen, ii, “ there is no person of a horse or ox or any
other of the animals which dumb and unreasoning live a life of sense alone, but we
say there is a person of a man, of God, or an angel.”
’ “ Norm,” pramapa, “ which is now as it ever was,” apurvddi ; >St. Augustine’s
“ Wisdom uncreate, the same now as it ever was, and the same to be for evermore ”
{Confessions, ix, 10) ; sandtana dhartna, the everlasting, self-revealed, and self-
consistent Veda. It is assumed that the Veda is not the work of any author, human
or divine, but presents itself to the divine omniscience in eternity (cf. the Christian
doctrine of “ eternal reasons ”), and has been revealed (cf. JJg Veda, x, 71, Miiniaka
Up., i, 1, etc.), for which reason it is commonly spoken of as ir-uti, “ that which has
been heard,” as, for example, in verse 6 below, where huti is rendered by “ self-
revealed word ”. For a fuller exposition of the fundamental doctrine of the eternity
of the Vedas see the MXmdrhsd Kydya PrakdSa, 6, in Edgerton’s version. New
Haven, 1929.
The doctrine does not, of course, imply that the actual words in which the Veda
is expressed are themselves eternal, but that that which was heard, and has been
transmitted, is in itself eternal. Christianity in the same way speaks of ” eternal
truths ” without implying that the ipsissima verba of Scripture in any one language
are eternal as such.
“ Being, Intellect, Beatitude,” sac-cit-dnanda ; cf. St. Thomas, Sum. Theol.,
i, q. 26, a. Ic, with respect to God’s beatitude, “ Beatitude is the perfect good of an
intellectual nature.”
* “ Of no-otherness,” advaya. The Vedanta cannot properly be defined as a
monistic system, but only by its own name as a doctrine of “ No-otherness ”
(advaita) ; “ God ” {tsvara) being spoken of as “ One ” rather in relation than as He
is essentially, viz. “ Without-otherness ”. *' No-otherness ” excludes both number and
quality ; “ That ” is inconnumerable and simple. This is also Christian doctrine
(Boethius, De Trin. ii, nulla iyitiir in eo dieersitas . . . nec numerus). The conclusion
follows that things in their thingishness {yathd-bhuta, i.e. as they are accidentally, or
to use a familiar Christian expression “ in their creaturehood ”) are non-essential
(an-dtmya), which is the Mahayana formulation ; or, in other words, that things have
essence (dtman) only in abstraction from the accidents by which we apperceive them,
which essence is therefore “ Not, not ” (neti, neti) anything that can be predicated
of the thing in its thingishness (astitva), and this is the Vedantic formulation.
* “ In the very principle,” cf. “ on the canvas of the Essence,” cited above in
note 3.
TWO VEDANTIC HYMNS
99
Tor ahodha and bodha we prefer to retain the primary senses of “ slumber ” or
“ unawareness ” on the one hand and “ awakening ” on the other, although the more
familiar “ ignorance ” and “ enlightenment ” are also implied. For in this way the
continuity of the traditional thought is better seen ; procedure from potentiality (the
condition of asuratia) to act (the condition of deiatva) being constantly described as
an awakening from stupor or slumber ; this is found especially in connection with
Agni as usarbudh “ awakening at dawn ”, and in connection with the Angels generally',
who are sometimes referred to as “ wake ", i.e. " quick ” in the Biblical sense. In the
same way the Buddha's “ enlightenment ” is really a “complete awakening”, samhodhi.
But life in the worlds, being stiU an admixture of potentiality and act, torpor and life,
is a continuous process of awakening, which when it is accomplished implies a
despiration (nirvdjia) and a return (niirtti) upstream (pratikula), and for this reason
the Complete Awakening is described in terms of “ inverse thinking ” (pratyak ceiana.
Yoga Sutra, i, 29) as “ Deep Sleep ” (see note 6), which Deep Sleep is contrasted with
the mere “ wakefulness ” (jagrat) of the empirical consciousness or possible intellect
(ahiddha manas, Maitri Vp.,vi, 34); attainment of the Ultimate Station (param padam),
that of the speculative or pure intellect (suddha manas or cit) implies a “ dementation ”
(amanihhnva, ib.). In other words, the whole course of the devayana leads from the
pre-rational, through reason, to the supra-rational.
“ Flux of things,” samsara, Eckhart’s “ storm of the world-flow ”. “ Uprush
of awareness of the Es.sence,” dtma-bodha-samunmesa.
** The maker of semblant worlds is knowledge-of or epistemological cognition
{avidya) as distinguished from knowledge-as or gno.sis {vidyd).
“ Self-revealed IVord,” sruti, as in note 7.
The doctrine of avindbhava, or “ necessary reciprocal relationship ", is here
enunciated ; see the many parallel passages cited in my A'etc Approach to the Vedas,
p. 9 and note 26.
“ Transmitted doctrine ” (disti) ; either with reference to the body of traditional
exegesis (smrti) or more likely with reference to oral instruction and initiation received
from a. guru.
“ Plenum,” purixam ; cf. Brhaddraixyaka I'p., v, 1 (almost identical with
Atharva Veda, x, 8, 29), “ Plenum from plenum taken, plenum yet remains.” This is
not a contradiction of avindhhdva (see previous note), for a plenum could not be
essentially infinite (but only numerically or relatively so) if anything remained external
to it ; i.e. “ without me ” it would not be infinite.
It is interesting to contrast this position with the difficulty that ensues when the
identity of divine and individual c.ssence D denied ; St. Thomas {Sum. Theol., iii,
q. 93, a. 1, ad 4) is forced to argue that “ .Although finite added to infinite docs not
make a greater thing, it makes more thing.s, since finite and infinite are two things,
while infinite taken by itself is one ” ; a tissue of logical inconsistencies, for it is evident
that an infinite to which a finite could be added cannot have been an infinite, but merely
an incalculable. Boethius knows much better when he says that “ Two belongs
to the class of things ” {duo rebus svut, De Trin., iii). Infinite plus infinite does not make
two things, because infinite is not a “ thing ”, cf. Bbhme, “ God is properly called no
thing,” cf. Erigena's “ God does not know what he is. because He is not any what ".
** “ That thou,” i.e. the well-known Aupanisada formula, “ That art thou ”
{tat tram asi), here contracted to “ That thou ” (tat tram).
P6st(a)
By fjARL ChaRPEXT[ER
“VrEITHER in Hobsoti-Jobson nor in Dalgado's excellent Glomirio
' Luso-Asidtico is found a single word concerning a word ])Of:t(d\
meaning an infusion or decoction of opium, which seems to have
plaved rather an important role in Mogul times. As several passages
ill the works of older European authors giv'e the word in the form post
(or at least somethmg very near to it) I shall here put together what
I have foimd about it. Probably several passages have so far escaped
my attention and may be added later on.
The oldest author in whose work I have met with this word is
Father Monserrate. Speaking of a certain Baba Kapur at Gwalior
he tells us as follows ^ : " Namque pmicis abhinc ainiis. m line radem
ciidtate. nebula quidem, cui nomen end Baba Capurins. Mabammeddi
sectator, fait, qiii Bacchi exoletos mores, et institnta poiione quadnm,
instmirauit , quae ex papaveris putnminihns aqua dihdis, co)ifici(ur.
Opinabatur homo perdilus, in eo beatitudincm esse positam, si omni
sensu quis careat, nec corporis morbos, nec animse eegritudincs patiatur,
quin magis, quodani sensuutn lenocinio semisopitus deliniatnr. Cnmque
id oppio offici animaduertisset, cui tamen si quis assuescat, in discrimen
incidit preematuree mortis, ex papauerum siliquis sorbiunculam exenqi-
tauit, in hunc moduni. Posteaquam oppium, ex papaueris siliquis
incisis effluxit, aeque maturitafeni sunt consequutx, demptis seminibus
in aquani conjiciuntur, tamdiuque subiguntur, donee humor quasi
vinum coloretur. Qui posteaquam paulisper resedit, in uas aliud reticulo
interjecto, tenuissimi Uni, transfunditur, el expurgatur, eiusque iam ad
potandum parati ipsi pharmacopai magnis crateribus se prolnnnt .. A
Xouus hie Epicurseorum conjphseus. plurimos nactus est, qui vi nom-en
dederunt, viros quidem principes, ct in ijs regem ipsum ZelaldinumA
htagnum- ^ .• qui omnes ab eo pharmaco, quod uulgari eoram lingua,
postum dicitur, Postinos^ se did honorijicum arbitrantur."
' Mun'juUriE Legationis Commentariu^, JIAttB., iii, ~>51 sq.
- Opium-drinkers abstain from meat, onion, garlic, etc., and especially from
oil, rvliich to them acts like poison. Owing to their continuous sleepiness they abstain
from sexual indulgence, and thus obtain a great fame for holiness.
^ .falalu-d din.
* Thus in the text, but probably to be written Mngrium as it is no doubt meant
to render Akbar.
® Posti, an opium-drinker.
102
JARL CHARPENTIER —
Post, according to Monserrate, is prepared from the husks of the
poppy steeped in water ; neither condiments nor spices are mentioned
here.
In another passage ^ Monserrate says of Akbar : ” Posti potione, uel
aqua sitim depelUt : cuius imniodico potu stupcfactiis sedet, et oscitalur.’’ ^
The next author known to me, who makes mention of pdst(d) is
Teixeira in chapter vi of his Relaciones de los Reyes de Persia,^ where
speaking of ajion (opium) he says ; ” Poor people use the husks, and
a decoction thereof instead of opium, and as the husks are called
past, those who do so go by the name of pustys." Pasty, of course,
is nothing but posti, though Teixeira seems to hav'e been thinking
of the Persian past vile, low
Teixeira is quoted by De Laet De imperio Mayni Moyolis (1631),
pp. 104 f., though the reference seems to be also to another passage
which I have unfortunately not been able to unearth. Says Dr. Laet :
■■ Opium preeterea, quo natmtes p>lerseque Asiatics tantopere capiuutnr,
maxima copia hie* coUiyitur ; ipsi meant Afion : cmanat illud e
surcuUs papnveris incisis : tenuiores ® ivro illius loco utuntur vnlyo
decocto capitellorum papaceris : quia autem rapitella estu vocantur
ipsis Past, eos, qui decocto illo utuntur, appellant Pasty, uti opulentiores,
qui opium xmirpant, Afonii. Usum illius primo introduction volant
Persee a viris principihus, quihus curce somnos adimebant : vulyusque
sensim imitare ccepissc. ut fere solent ; ita at jam pauci sint, qui
alterutro non utuntur. Narrat Teixeira se ridisse nonnullos. qui
tantopere illi assuecerant. ut quum illo de.stituerentur, emorerrntur :
plures tamen niniio illius usu laeduntur, atque adeo tolluntur, est enim
lethale venenum, nisi quis modum adhibeat. d; caute illo utatur."
The Jesuit Father Bartoli in his Missione al Gran Mayor did Padre
Ridolfo Acquaviva (1663).® p. 64, tells us the following about Akbar ;
“ 0 se pur I'invitava a dirgli alcuna cosa di Dio, appena cominciato,
s'addorrncntava ; e cio per lo troppo uso hor dell'Orraca, che e
un fumosissimo vino di palma,’ hor del Posto, che e una
‘ MASB., iii, 642.
^ Cf. Smith, Akbar, p. .336.
^ Tran.slated in the Hakluyt Society Puhl. (1902), p. 2O0.
‘ Viz. in Gujarat.
5 These are the “poor |)eople ” of Teixeira.
® There are other edition.s issued at Rome in 1711 and at I’iaeenza in 1S19.
1 There is considerable confusion in the use of the word Arrrirk. Hark (rf. Hob-mn-
Johson, S.V.). It is undoubtedly mueh u.sed to denote the spirit di.stilled from the
exudation of the date-palm, which, however, .should pro{H>rIy lie called todity ; hut
it also means a coarse brandy distilled from cane-mola.s.ses and especially from rice.
p6st(a)
103
r'
■?
-I
c
E«
■s
tal confettioae d’oppio, rintuzzato, e domo con varie correttioni
d'aromate."’ Bartoli no doubt is dependent upon Monserrate whom
he has even slightly misunderstood ; but the mention of spicing the
post, which appears here for the first time, must have been taken
over from another authority.
The famous French physician Fran9ois Bernier spent at least ten
years (1658/59-1668) within the frontiers of the Mogul Empire, and
obtained a thorough knowledge of things Indian. In his work there
is found the following passage concerning p5st{d) ^ : “ Ce Poust
n est autre chose que du pavot ecrase qii'on laisse la nuit
tremper dans de I'eau ; c'est ce qu'on fait ordinairement boire a
Goiialeor, a ces princes auxquels on ne veut pas faire couper la tete ;
c’est la premiere chose qu'on leur porte le matin, & on ne leur donne
point a manger qu’ils n’en aient bu uue grande tasse. on les laisseroit
plutot mourir de faim ; cela les fait devenir maigres & mourir insensible-
ment, perdant pen a peu les forces et I'entendement. & devenant
comme tout endormis et etourdis, & c'est par la qu'on dit qu'on s'est
defait de Sepe-Chekouh,^ du petit fils de Morad-Bakche, & de Soliman-
Chekouh'® meme.” The chief state prison of the Moguls was at Gwalior ;
and Bernier seems to be the first European to mention the custom of
slowly poisoning State prisoners to death with pdsl(d),* which was
probably practised during the time of Shah Jahan (1627-1658) and
certainly during the earlier years of the reign of Aurungzeb
(1658-l'707).5
The native historian Muhammad Stilih Kambii in his ’Amal-i
Salih (v. Elliot-Dowson, History of India, vii, 131) also tells us that
Sulaiman Shikoh and iluhamraad Sultan (a younger son of Aurungzeb)
were taken to Gwiilior to be fed upon kukndr, which is simply a
synonym of pdst{d). Muhammad Sultan afterwards was removed to
Sallmgarh and there done to death by opium.
‘ I quote from the ettition of .Amsterdam, 1724, called the Histoire des Elats du
Grand 3Iogol, p. 147. An English translation is found in the Travels in, the Mogul
Eynpire, by Constable and Smith (1916), pp. 106 f.
“ .A younger son of Dara Shikoh.
* The eldest son of Uara Shikoh.
* Cf. however, De Laet, l.c., p. 40, who, speaking of the State prison at Rantham-
bhor, tells us that prisoners were kept there for two months and then : praefectus
arris eos producif, el in fastigio muri conslitutos et lade potatos, praecipites agit in
suhjedas rapes. The lac mentioned here must certainly be some sort of decoction
of opium (or possibly bhang which is sometimes mixed with milk, cf. Watt, Commercial
Products of India, p. 258).
‘ That such was the case is mentioned by Tavernier, Travels la India (ed. Crooke),
i, 52.
104
JARL CHARPENTIER
Peter Mundy in his Travels, ii, 247, also mentions the p6st{a) ;
“ Of the hushes they make a kind of Beveredge called Post, steeping
them into water a while and squeezinge and strayninge out the liquor,
they drinck it, which doth inebriate.” Drinkers of this decoction
are called Postee.
A curious deviation from all other descriptions of this beverage
is found in the New Account of East India and Persia, beimj Nine
Year's Travels, 1672-1681, by the well-known English physician
John Fryer ^ (1650-1733). In i, 92, we read the following notice
concerning the punishment of great nobles within the Mogul's realm :
“ Upon an Offence they are sent by the King's Order, and committed
to a place called the Post (from the Punishment inflicted), where
the Master of the Post is acquainted with tlie heinousness of the
Crime ; which being understood, he heightens lyv a Drink which at
first they refuse, made of Bamj (the juice of the into.xicating sort of
Hemp), and being mi.ved with Dairy (the deadliest sort of Solannni or
Nightshade) named Post, after a week's taking they crave more than
ever they nauseated . . . making them foolislily mad '’ etc. The
same recipe for preparing Post is repeated i, 263 ; in iii, 169, Fryer
mentions “ Goualar or Post ", apparently believing Post to be a name
of the state prison at Gwalior.
It is scarcely probable that Fryer, who was himself an M.D. and
apparently much interested in natural history should have been so
badly informed that he thought post to be not an infusion of poppy-
husks steeped in water, but bhang (prepared from Cannabis) mixed
with Daturad But if we take it that in the passage quoted above
' Edited for the Hakluyt Society by \V. Crooke, vols. i-iii, London, 1009-191.5.
The tirst edition appeared in lOOS.
- A few references to Dutnin may not be out of jilaco here. The species, whii h are
put to medicinal and especially criminal use in India seem to be above all /h/uv/imsn h.
(; D. nlha Nees) and I). Met'-l L. It is know ri to the Portuguese authors as Diitio ( Konkani
dhutro) or Dut/oii {in Port, ami .Spanish also as hnrlndiiiii " joker ” because it makes people
laugh in a foolish way) ; .Sanskrit ilietioiiaries usually give words hke dlnilfuin or
dhuftunt, the connections of wliiih are by no means clear (a suggested relationship
with Latin /t.s/ii. n - a straw ” cannot be uiiheld as there is scant reason fur believiu”-
dhattura. dhuitunt. etc. to be of .Aryan origin). IhUnKt seems to be mentioned first
of all by Garcia d'Oita, CoUnr/ttio.t. x\ ( I.56:{) : “ .Sekva. A' niinh.i senhora dcii dnturn
a beber huma ncgra da easa, e tomoiilhe as diave.s. e as joyas quo tinha ao [lesco^o,
e as qiie tinha na caixa. e fogio, com outro negro . . . • Orta. .V queni dam esta mesinh.i
nao falam cousa a proposito ; e seinprc riem, e sao miiito liberaes, c todo o ncgocioo
rir e falar niiiito pouco, e nao a proposito ; e a mancira que qua ha de rouliar he dei-
tandolhe esta mesinlia no comer, porque os faz estar com este acidente vinte e qiiatro
horas.” llonserrate, l.c., p. 574, speaking about the institution of suttee tells us how
the poor widow-, before being burnt, was stupelied by poison : “ //ns mulierculas. A
post(a)
105
from De Laet, p. 40, lac really means hhang diluted in milk there
certainly also exists such a drink with an admixture of Z)«0(r«-seeds,
which is considered to be extremely dangerous and maddening to
any degree, cp. Watt, l.c., p. 488. It may be very well possible that
State prisoners were done to death not only by preparations of opium,
but also by other poisonous concoctions, which were all popularly
known by the common name of post.
Malcolm Mem. of Central India. 2, ii. 146 n.. mentions the licpiid
opium which he calls not po-d but kusoombah.^ It was much in use
as a ceremonial drmk with the Rajputs and considered as a sacred
pledge of friendship. Such a use of various infusions of poppv is
also mentioned by Tod,^ who also docs not seem to use the word po.d.
Post is known also to Burton Scindc or the Uidnt ppij TaVcy. i, 267 sq.,
where he describes it thus : '' A dried poppy-liead or two was infused
in warm water allowed to stand the whoh' night, and in tin* moriung
squeezed till none of the juice remains in it." The drauglit was
then cooled in ice or snow in the hot weatlu^r. sweetened, perfumed,
and thus administered to the captive.
Other references that I have come across are to Tlievenot. TrnrcU
(1687), ii, 97 ; Hamilton, .4 Sew Account of the East Indie.s. i, 172. and
Herklots' Qanun-i Islam, p. 326. but as they add but little to our
knowledge of pdst{d), I leave them out here. A reference to Forbes,
Rds-Mdla, p. 557. must probably be incorrect ; at least it has proved
impossible to me to identify it in the new edition of that work which
is alone accessible here.
Post or posta, according to the authorities quoted above (with the
exception of Fryer), thus is an infusion of poppydieads steeped in
omni (lolnriA ^en.-^u careaiif, qnthuHfahi^opjiio fjra'ttitim. rdhciha tiopni
{quae banijiie dicitur, cauafdquc quam es/) oily quad t.sf, dnlurone hihn,
Iudi'< I'ognita, Euwpaci--^. ac veftrih>(< proiAv.^ iqtadn ommun." It is spoken
of ajs havini; been mueh used in Goa to stupity jealous l)U'>ban(l> in order to affonl their
wives more freedom, tf. e.i;. Linx hoten, Vtoja/p On K. ItiditA, i, ; Pyr.ird de
Laval, yoija'jp {Hakl. Soe.), li. Moeijuet, Toyt/j/r'-, p. 31J, etc. That it was
used by Indian rtibbers to stujiefy their vietiius was known already to I’ru>per Alpinu’^,
Hist. Aerjijpt, (1580), i, 190 sq.: and it i?* ^tiU ^.aid to l>e in frecpient u^c with tlie
descendants of thugs, cf. e.g. Slecman. Ramhhs and ReeoUntuni't (ed. Smitli), p. 82 sqq. ;
ChevtTs, Ind. Med. J urif<prudi)n'€y p. I79sq<j. Watt, l.c,, p. 488. tells us tliat pots in
which arrak is jiourcd are sometimes fume<l with the smoke of burnt Ihitura-^vay.
On iJatu/a, (f. further Acosta. Trartndo de la Rnuja-"', p. 87 ; Kheetle, Mala-
baricu.s, ii, 47 sqq. ; Kumphiu'J, Herb. Anibtnnen''e. v, 246 ^qq. ; Ainslie, Hai. Mai.
Hind., p. 47, etc.
^ 8’kt. Ku.sut/ibha is the Safflow'cr or BastanI Saffron {< 'urthamu^ L.) ;
a corre.s|)onding word also seems to he U'.ed of the Lae Tree {Sddeichera trijuja Wilhl.).
2 Cf. Rdjadhdn {ed. Crooke). i. 86, etc.
106
p6st(a)
warm water and allowed to draw over night. The fluid is then strained,
sometimes cooled, and perhaps mostly mixed with spices or other
flavouring stuffs. So far there seems to be little difficulty. That it
should be in special use in the Punjab ^ is, of course, quite possible,
though it seems to be widely known in different parts of India.
There remains, however, the derivation of the word post
or postd (tprw) itself, which seems to cause some uncertainty.
Sir George Grierson, whose knowledge of everything connected with
India is unrivalled, and who was for a couple of years himself an
Opium Agent in Bihar, tells us - that the usual name of the Papnver
somniferum L. is really post or postd, which thus, at least within
Bihar, denotes the whole plant. I ventured to write to Sir George
Grierson on the subject, stating it as my humble opinion that it is
really the seed-capsule that is called post{d), and that this name
was then transferred to the plant itself ; and in his reply ® Sir George
willingly endorsed this opinion. He further tells me that the
common opinion in Bihar seems to be that post is really the Persian
word post meaning “ skin which is “ in this connexion
referred to the ‘ skin ’ or outer shell of the poppy capsule For
phonetic reasons it seems impossible that pdsl{d) could be a genuine
Indian word, and it must thus be suggested that it was borrowed
from somewhere. The difficulty .seems to me to be that Pers. pdst
does really mean “ skin, hide of an animal and that it seems slightly
uncertain whether such a meaning could be developed into the “ skin ”
(or rather shell) of a poppy capsule. As, however, no other probable
derivation seems to be at hand we shall so far have to rest content
with this one.
^ Watt, I.C., p, 84.").
“ Hihar Pea'^nnt Life, 2nd od., p. 241.
^ Letter dated 11th January, 103J.
^ On this word which has been borrowed into Sanskrit as pa3tn{lfi)- manuscript,
book " cf. Gauthiot, MSL., xix, 130 f.
^ Pers. post should be related to Kurd, pl^t “■ skin ” {G.I.Ph., i, 2, 267) and to
Avestan pav/a- skin There is considerable difficulty concerning the root-vowel ;
but undoubtedly the Avestan word owes its origin to a false writing and should really
be piv-3t or p'n"<t (i.e. *pu'ita- or *pftu-^ta-).
Bhagavata Parana and the Karikas of Gaudapada
By Amarxath Ray.
i BOUT three years ago, I sent a paper on “ The Date of the Bhaga-
vata Purana to the I.H.Q. The publication of the paper \yas
delayed, and it was forestalled by B. X. Krishnamurti Sarnia's paper
on the same subject, which appeared in the A)nMh of the Bhandarkar
Oriental Research Institute, vol. xiv, pts. iii-iv. The object of both
the papers was the same, viz. to controvert the views of Vaidya and
Winternitz who proposed the tenth century a.d. as the date of the Bh.P.
Sarnia suggests that this Purana was composed in the fifth century,
if not earlier, ily own view is that the work came into lieing some
time between a.d. 550 and 650. The mention therein of the Huns
(ii, 7, 26) and of the Tamil Saints (.\i. 5. 38-40) would go against
Sarma’s h_\’pothesis. Sarina and the present writer adopted somewhat
different lines of attack upon the position taken up by 'S’aidya and
Winternitz. It is unnecessary, however, to state the additional
matter my paper contained, or to publish it. This will be done if the
other view finds a defender who lias to be refuted.
I propose in this place to raise a discussion as to whether the
composition of the Bh.P. preceded or followed that of the Kiirikas
of Gaudapada. From the parallelisms cited below, it will be clear
to all that one of these two imters must have been influenced by the
other. Sarma points out a quotation from Bh.P. (.x, 14, 4) in the vrtti
on the “Uttaragltii” (ii, 45). attributed to Gaudapada. and also two
clear references to the Bhagavata in the so-called Mutham vrtti on
the “ Samkhyakarika ". As regards thus latter work, it has been
doubted whether we have the original text before us ; the work
appears to have grown with the times, and the Bh.P. references
found therein do not appear in Paramartha's Chinese translation.
If the Uttaragita commentary were a genuine work of Gaudapada,
as Belvalkar and Sarma think, it would be clear that the author
of the Bh.P. preceded Gaudapada. I must, however, invite the.se
and other scholars to reconsider the question in the light of what
follows.
Compare the similarity of ideas in the following quotations from
the Bh.P. and the Karikas of Gaudapada : —
108
AMARXATH RAY —
(а) “ Addvante ca yan ncisti vartamane' pi tat tatJid
Yitathaih sadrsdli santo vitathd iva lahptdh.'’
G. K., ii. 6, and iv, 31.
“ Xa yad idam agra dsa tia bhavisyad ato nidhandd-
Aniimitam antard tvayi vibhdti tnrmihirase
A ta upa ml gate d ra vi najdt i vikaipnpathn ir
V itathamanovildsam rtam ityavaynntyabadhdh
Bh.P., X, 87. 37.
Adyantayer yad asatdsti tad era madhije.’’
Bh.P., xi, 19, 7.
“ Xa yat purastdd uta tan na pascdt
Madhyejn tan na ryapadesatndtram.''
Bh.P,, xi. 28, 21.
Note the word vitatJia in the Karikii and in the first of the Bhagavata
passages.
(б) Xa >tirodho na cetpatti na baddJio na ca mdhakah
Xa mumakpir na rai makta itg em paramdrthatd."
G. K., ii. 32.
“ Baddho mukta iti cydkhyd gunate me na va-dutah
Gunasya mdydnadatcdn na me mokm na bandhanam
Bh.P.. xi. 11, 1.
(c) ■■ Advaita)H -samanuprapya jadavallokam-dcaretX
G. K., ii, 35-6.
■■ Atmdrdmd nayarrttyd vicarej jadaranmunili."
Bh.P., xi. 11. 16-17.
{d) “ Mdydisa tasya devasya yayd yani mohitah svayamX
G. K., ii, 19.
“ Sramdydgunam. drisga bddhyabddhakatdni gafaJiA
Bh.P., vii. i, 6.
(e) " Ghatddim prallncmi ghatdkdmdayo yathd
Akdse sampndlyante tadraj jlvu iJidtmani."
G. K., iii, 4.
" Gliate hldnne ghatdkdsa dkdsa sgdd yathd purd
Eratn dehe rnrte jlro brahma sampadyate punah.’’
Bh.P., xii, 5, 5.
(/) '■ Manodriy((m idani draitam yat kinrit sacardcaram
Manase hy amanlbhdre dvaitatn nuicopalabhyate.”
G. K., iii, 31.
BHAGAVATA pueana axd the karikas of gaudapada 109
“ Manah srjati vat dehdn gundn karmdni cdtmanah
Tan manah srjate mdtjd tato jivasya samsrtih.”
Bli.P., xii, 5, 6.
(g) In interpreting Bh.P., xi, 24. 7, Sridhara quotes tlie well-known
G.K., iii, 15 : —
“ Mrllohavisphulingddyaih srstiryd coditdnyathd
Updyah se’vatdrdya ndsti bhedah kathancanaT
(h) The similes of the Rope and the Snake and of the city of
Gandharvas, so familiar to Advaita Vedanta, occur in both the works.
(Bh.P., vi, 9, 37 ; xi. 26, 17 ; and G.K.. ii, 17-18 ; and Bh.P., vi,
15, 23, and G.K., ii, 31.)
(i) Bh.P. (\di, 15, 54) names the four states or modes of the Being.
\dz. i-jsra. taijasa, prdjna. and Inrya, just as they occur in the Karikas.
which Vidhusekhara Bhattacharya considers to be older than the
so-called Mandukya Up., where the first and the fourth appear as
Vaisvanara and turlya.
A closer comparison of the two works would, I am sure, disclose
many more parallelisms. In the absence of any earlier semi-Buddhistic
Vedanta work of the tcpe of these Karikas, a student of Indian
Philosophy is naturally led to think that the Karikas were earlier
than the Bhagavata Purana, which looks like attempting to harmonize
their Vedanta with the Pancaratra Bhakti religion, without, at the
same time, disowning the latter's original allegiance to what is known
as the Pauranic Samkhya. But while the author of the Purana appears
to be familiar with the “ Ajatavada doctrine, he does not know the
“ Anirvacanlyata ” doctrine posited by Sankara. I am inclined,
on this among other groimds, to think that the author came between
Gaudapada and Sankara, No doubt the latter does not mention
the Bhagavata in his commentaries, but if the “ Govindastakam be a
genuine work of his, as the sixteenth-century Bengal Vai.snava ^\-riter.
Jiva Gosvamin thought, and as Belvalkar and others of tliis day
think, Sankara must have known the Bhagavata. It may be noted
also that both Sankara and Ramanuja are said to have known Puri
or Jagannath, but that holy place does not appear to have been known
to the author of the Bh.P. It would otherwise have found mention
in the chapter on Balarama’s pilgrimage (x, 79).
The difficulty in the way of the acceptance of my h^-pothesis is
twofold : Firstly, the Bhflgavata passage, hunted out by B. N. Krishna-
murti Sarma, from the " Uttaragltavrtti,” and, secondly, the tradition
that Gaudapada was the teacher's teacher of Sankara.
110
AMARNATH RAY
The first difficulty is not really so insuperable as it appears. The
attribution of the vrtli to Gaudapada rests on the evidence of the
colophon in one or two manuscripts of the text. Such colophons
cannot form independent evidence, knowing, as we do, that they
often originated from ignorance or fraud. Xor does the fact that
the philosophy of the vrtti is Advaita of the Gaudapada type mean
much, for that is also the philosophy of the ‘"Uttaragita" itself. I should
be prepared to ascribe that work, rather than the commentary, to
the great Gaudapada.
The second' difficulty is harder to meet, as in meeting it, one has
to reject a time-honoured tradition. The traditional relation between
Gaudapada and Sankara would not leave sufficient time for the
pre-Sankara author of the Bh.P., belonging to the extreme south of
India, to be famiUar with the Karikas of Gaudapada. Two pieces
of evidence are adduced in support of the tradition, viz. the mention
of Gaudapada as paramaguru in the puspikd to the commentary
on the Gaudapada Karikas, usually attributed to Sankara ; and the
reference to Gau<Japada as guror ganyase in the metrical “ Upadesa-
sahasri” (xviii, 2), an undoubted work of Sankara. But can either
of the expressions mean nothing else but teacher's teacher ” ? I see
no reason to restrict the meaning in this way, in either case, though
paratnaguru is generally understood to mean “ teacher’s teacher ”.
I have besides grave doubts as to the authorship of the commentary
on the Karikas. Vidhusekhara Bhattacharya would reckon it among
the Sahkaracarya apocrypha,^ and I agree with him for a reason
which he does not adduce, viz. the author’s ignorance of the real
import of a good many Buddhistic philo.sophical terms which abound
in the Karikas. Sankara was more familiar with Buddhism than
any other commentator on the Brahmastitras we know of, and I
am not prepared to ascribe such ignorance to him. One has only
to consider the widely varying attitudes of Gaudapada and Sankara
towards Theism, as also towards the Buddha and Buddhism, to be
convinced that more than one teacher must have come between the
two.^ I do not propose to enter into a detailed discussion of all that
has been said by others about Gaudapada and his age, but I am
inclined to agree with Barnett and Jacobi in thinking that a date,
' Sir Atthitosh Silver Jubilee Memorial Volume, iii, pt. ii, Sankara’s commentaries
on the Upanisads !
2 Xote also the remote nature of Sankara’s references to Gaudapada in his com-
mentary on the Brahma-Sutras, i, 4, 14-5, and ii, 1, 9), as “ Sampradayavit ” and
“ Vedantartha-sampradayavit ”.
BHAGAVATA PUKANA AND THE KARIKAS OF GAUDAPADA 111
later than a.d. 500, should not he assigned to this great teacherd
This would place him nearly 300 years before Sankara, and allow
for an interval, during which the Karikas might travel down to the
south. May it be that the influence of the Purana itself led Sankara
to find a respectable place for a personal God and his worship in his
system ? Gaudapada appears to have felt little concern for Theism
and bhaJcti.
^ See Barnett’s review of Walleser’s work in JBAS. 1910, and Jacobi’s paper on
“ Mayavada ” in JAOS. 1913.
Some Early Dramas in Bengali
By Jayanta Kumar Dasgupta
T) AMNAEA YAK'S Kulin Kulsarbasva (1854) is often regarded as
the first Bengali drama, but there are many other Bengali
plays of one sort or another anterior to it in date.
Although Herasim Lebedeff, a Russian adventurer, staged two
Bengali plays in Calcutta towards the end of the eighteenth century,
these have not come down to us. It is doubtful if Lebedeff's plays,
which were translations of Enghsh dramatic works, were ever
pubhshed. No clue to the subsequent fate of these plays is found
in the autobiographical introduction to Lebedeff’s Grammar of Pure
and Mixed East Indian Languages, printed in London in 1801.
That some kind of indigenous Bengali plays existed in the first
part of the nineteenth century is evident from a reference to an
old play named Kalirdjar Ydtrd in some of the old Calcutta periodicals.^
This play seems to have been popular in its own day. The characters
are two Vaisnavas, Kaliraja, his Minister, his Guru, a noble and well-
dressed EngUshman with his lady and two servants. Dancing, singing,
and witty conversation were distinguishing features of this play,
which was very much liked by the audience.^
The tone of Bengah plays in those days was rather low. Dramatic
versions of Vidyasimdar were severely criticized by contemporary
journals on this ground.^ In 1826, the Samacar Candrika put forward
a vigorous plea for a regular stage.'* For some years before theatres
were founded in Bengal, translations Mere made from Sanskrit to
supply the want of good dramas in Bengali. Some of these earlier
plays M'ere until recently very little known.
A paraphrase of Krsnamisra’s Prabodha Candrodaya NdtaJca,
M-hich was published in 1822, is the first printed Bengali drama. This
paraphrase under the title of At matall cakaumudi was the joint work
of three pandits — Kasinath Tarkapancanan, Gangadhar Nyayaratna,
1 Calcutta Review, vol. xiii, p. 160. The Bibliotheca Orientnlis, vol. ii, p. 460,
mentions a Bengali work, Knliraj, but it is not definitely stated whether this was a
drama or a prose narrative.
Asiatic Journal, 1822, p. 287.
’ Calcutta Journal, 26th February, 1822, p. 587.
* Asiatic Journal, 1826, p. 214.
VOL. Vin. PART 1.
8
114
JAYANTA KUMAR DASGUPTA —
and Eanikinkar Siiomani. The whole composition is in a very stiff
kind of Bengali prose. In 1830 the Saniacar Candrika advertised a
version of the same play written in '■ payar ” verse, but nothing furtlier
is known about it.^
Janies Long's Granthabali or Catalogue of Eleven Hundred Bengali
Books (1852) mentions another drama. Kautuksarbasva Xdtak.
This is based on Gopinath Cakravartl's Sanskrit play noticed by
Wilson in the Select Specimens of the Theatre of the Hindus. The work
was done by Pandit Ramcandra Tarkalariikar of Harinabhi. A copy
of it which is in the British Museum is dated 1828. Some people have
confused this play with a stage-version of V idgasundar performed
in Calcutta in the early thirties of the last century.^ Kautuksarbasva
is a two-act play and opens with an invocation to the god Ganesa
in “ tripadi ’’ verse. The characters are Kalibatsal Raja, his general
Samarjambuk, a learned Brahman, Satyacaryya, some courtiers,
the Queen, a maid-servant, and a public woman. It shows occasion-
ally a bad moral taste, being a play which aims at illustrating the
degenerating tendencies of the Kali age. The style is a Tui.xture of
the Sanskritic and the colloc|uial. The translator, however, regarded
his language as ” sadhubhasa ”. The Bengali version is truly speaking
a paraphrase rather than a translation of the Sanskrit original. Both
prose and ver.se are u.sed and there are many stanzas in " payar
and ■■ tripadi
Long's Catalogue of the Vernacular Literature Cominittec's Library
(1855) names a play Mahanatak, Rani's History dramatized from
the Sanskrit, 1840. This e\ddently is the .same play as Ramgati
Kabiratna’s dramatized ver-ion of Mahdndlaka. the date of which is
given 1)}' Schulyer as 1849.® This is not a regular drama. It is a
kind of dramatic represemtatiou of Rama's life. Th(‘ j)lay opens with
an invocation to Ganesa. " Payar and other forms of Bengali vcr.sc
have been used throughout. There arc also several stage-directions.
Parts of this play do not evince good taste even if some
allowance is made for conv(>ntional >Sanskrit erotic descriptions.
Another translation from Sanskrit, Jaga<llsvara's Hdsydrnava, is
supposed to have been publi.slnsl in 1840.* The British Museum copy
^ Samficar Candrika, IS.SO, 12th April.
2 Dhnitnhjtin Mukhfipadhyfuj-BayHfujn Xfiti/ftAtlld, p. 2.
^ A Bihhoqraphy of the Sfia^krit Dtninii, ]>. Ha. Lori^, houTvrr, in his l>K''rtptirp.
Catalogiif of BentjaU H'orA’s (1855), gives the date as IS-th.
^ Bendall. CataJogiie of Pali, Prakrit, amt SiDi'^knf ]\'ork-< irt (hr Prifi^h
(189^), pp. 143-4. The Jiihliothfnt t)rit‘ntah'<, vol. li. p. 4.">4, gi\cs th<- date as 18,3.5.
SOME EARLY DRAMAS IX BENGALI
115
of this play bears no date. Wilson noticed this play also. The taste
of Hdsydrmva is at times absolutely indecent and the humour is
spoiled by blunt coarseness. Wilson says that the King left at the
end of the first act. In the Bengali version there is no stage-direction
to that effect, nor is there any division of the play into regular acts.
One has to assume that the King left with his whole partv as the
latter part of the play is monopohzed by the pandits and the courtesans.
The narrative portions of Hdsijdrnav<t are in Bengali prose and transla-
tions of the original Sanskrit slokas are in “ payar ".
A Bengali translation of Sakiintnld by Ramtarak Bhattacaryva,
a student of the Calcutta Sanskrit College, was reviewed liy the Bengali
periodical Sambdd Prahhdkar in 1848. But no trace of it has been
found so far.
A Bengali rendering of Srlharsa's RatadvaU by Xllmani Pal
was published in 1849. This play opens with a prayer to CTane.sa in
“ payar ". Then follows a sort of " bandana " or salutation to the
Guru, after which comes the ■' granthasucana or introduction in
“ tripadi ”. But the work of Xllniani Pal is not an e.xact translation.
It is only an adaptation. Nllmani Pal supplemented Srlharsa’s text
with additions and alterations of his own. In the earlier part of the
play, he introduced a description of Srlharsa's capital which is not
found in the original. Then there are other additions, e.g. a whole
story about Ratnavall and a description of a voyage. Almost all the
stage-directions of the original are omitted, though the division of
the play into acts remains the same as in Sriharsa. The dialogue
parts of the plav have been considerably condensed, being put often
in narrative form. The poetry is not altogether bad. Various forms
of Bengali verse such as payar. laghutripadT. ckabali, tripadi, dlrgha-
payar. ekahall antayamak. tuuakabhas, lalitlaghu. totak, and
caupadi have been u.sed. But ‘‘tripadi " and ” payar " seem to
have been the favourites. Nllmaiii Pal must have been a devotee of
the godfless Kali as there are several references to his reverence for
that deity in this play. There is more use of prose in the fourth act
than in other acts, and this part of the play, strictly speaking, becomes
a mere narrative ivith occasional stanzas of poetry. Therefore towards
the end there is a sense of monotony. ‘
^ Loii 2 , in his Dc^criptiif Caldloijur of lU ugnli If (18.').')) mentions this work
twice, ami in one place he remarks that it retpiires pruning. This Catalogue, which
is rather scarce now, has been reprinted as an appendix to Dr. Dinescandra Sen’s
Bangnhh(t.'ia o Saliitya (fifth edition).
Iranian Studies V
By H. W. Bailey
T he task of publishing the Khotanese Saka texts of the India
Office and the British Museum, on which I am engaged, is likely
to be long protracted. It therefore seems ad\dsable to make known
as soon as possible a list of the words for which parallel texts guarantee
the meanings in Tibetan, Chinese, or Sanskrit. Most of the following
words are from the Siddhasara-sastra of Ravigupta, of which there
are extant sixty-five folios. The Tibetan is to be found in the Tanjur.
Two short passages of the Sanskrit are preserved in a compilation in
a Madras MS., of which a copy (made in 1902) is in the Bibliotheque
Nationale. Through the courtesy of the Librarian I have been able
to make use of this.’^
M'^hen the meaning of the Saka is imcertain, the English equivalent
is given for the Tibetan parallel word, and follows it. An asterisk
denotes a translation from the context, without direct parallel text.
A large number of other words are known to which no meaning can
yet be assigned. Most of the words are new, but some, already known
in E, can now be certainly translated, others are given as corroborative
evidence to meanings known before. The abbreviation E refers to
E. Leumaim, Das nordarische (sakische) Lehrgedicht des Buddhismus,
1933-4 ; Konow Suv. to Sten Konow, Zwolf Blatter einer Handschrift
des SuvarnabhdsasiUra in Khotan-Sakisck, 1935 ; Sacu Doc. refers to
the Saka Text edited by Konow in Two Medieval Documents from
Tun-huang.
a^a
. flour
(NPers. ard).
. phye. See arrana.
adffiye
. other
. gzan ; adarasta to
another place ; ttalanasta
du. Cf. adana.
u adarasta to and fro phan ts'un
adlsta
unripe
. ma smin-pa.
agane
. potcerless
Sogd. yn’ Vimal. 97).
dbah-med-par (cf.
aha
noose
. Skt. pasa.
' Since the above was sent to the press, I have been able to use in Oxford photo-
graphs of two Nepalese MSS. of the Sanskrit text of the Siddhasara. It has been
possible to use this rich new material to a small extent during the reading of the
proofs.
118 H. W. BAILEY —
aharsta
regularly
rgyun-tu.
ahauvam .
U'eah
see hauvana.
ahvana
to he warmed .
bsros-pa.
ahvarai/em. ahvarrja .
sour
skimr-ba. Konow
Saka Studies ahvarrai.
amasta nye
unfermented curds .
zo ma lans-pa.
anesta vata
without ji rmness
Skt. asaraka.
See esta-.
amgausdi, amgusda
asafoetida
sin kun (XPers.
anguzad).
anvasta
difficult .
dkah-ba. Skt.
krcchra. See
nvasta. iMisread *atvasta- in Aparimitajuili
Sutra.
anuta
unaccustoned .
snon ma goms-pa.
Cf. nuska.
arraje
shrinking
bkbums-pa, see ari-
sama.
arrana, arana
*to be ground .
Ptc. arda ( < *ar-n-ta
cf. IskasmI yurd (/round).
arise bausa
bad smell
mnam-pa.
arisai
unpleasant
mi zim-pa.
arisama
shrinking
hkhums-pa. Skt.
.samkoca. Cf.
parisame decrease, and
arraje.
asana
green, blue
snon-po.
asarramata
dread zum-pa
Cf. E 8, 24, ssarri.
E 21, 8, assifda
asernja utca
la,ke water
mts‘ohi chu, Skt.
sarasa. Cf. E 21, 4, assimgye.
asnai
pigeon .
Skt. kapota.
askii . . . .
deer
askii Iliya gusta, Skt.
aina (sc. ma
msa) deer's flesh (Sogd
. asuk). So translate
E 10, 8, o tarra rrau viita aska or thirst]/ deer in the plain
(see rrau below).
askaphai .
poppy .
le brgan.
askhauysa
top
thor-to. Cf. kliauys-.
asnai
precious
thog-mar.
aspasdaka .
producing
aspasde, 3 sg. rgyur
hg}Tir-ro.
asthajana .
1. to he applied
bsgo zin.
2. to be taken
blaii-bar bva-lio.
IRANIAN STUDIES V
119
isthanijafia . . to be taken . . blaiis-la, blaii zin.
Pret. usthlyanda E 6, 30 ; 24, 160.
astaucii . . . dry land . . .skam-sa.
asurai . . , vnclean . . . See siirai.
atcasta . . . injured . . gnod-pa.
avyaucii . . . unbearable . . mi bzad-pa.
aysbrljsanii . . to be roasted, fried . briios-pa. See brrljs-.
(ays- iys- eys- are written for older uvs-.)
aysdau, aysdo . . boy . . . bvis-pa.
u vara ana ksasii salii vl bure aysdo gtirste and thence up to
sixteen years he is called aysdo.
aysdemafia, aysdimana to be cooled . . bsgrans-te.
aysdraphai . . crouching . . tsog-tsog-piir.
ajTila . . . ball.
pasve a^Tila . heated ball.
Chinese X in the story of Mahaprabhosa. So read Sacu
Document 70 ; ysiri bimdi brrlvlnai ttaudi ayfda on the heart
the hot ball of love.
ba, baka, bakalaka,
bata . . . small . . . niiii.
bajam, loc. sg. bajinana vessel . . . snod.
bajevakii . . . destroyer . . Skt. pranasaka.
bajsiha, loc. sg. mortar . . . gtun, E 5, 65, baj-
bajsihana. siha.
bale . . , chu skyar ( = Skt. a ivater-bird
kalaharnsa)
balgana, loc. sg. . vessel . . . Skt. bhanda.
baloha, loc. sg. balo-
hana . . .a cloth . . . ras.
balsa, gen. plur. balsanu stupa . . . mchod-rten.
In Sacu Document 42 basi paste padaide he ordered to build
a stupa.
bamame . . . vomiting . . sk}-ug-pa.
bamanahii . . . to be made to vomit . 3 sg. banie vomits.
bamii . . . vomiting.
banate . . . jack-fruit tree. . Skt. panasa.
bara . . . *crops . . (Mid. AVest Iran.,
NPers. bar.)
darabhaksa pasta u bara pahi * famine occurred and crops
failed.
bata, bava
root
rtsa-ba.
120
H. W. BAILEY —
beramda, . • craching . . hgas-pa.
bi ... tvillow . . . Ican-ma (Av. vaeti,
Soi vi, NPers. bid).
bijautta . • • spoiled . . . Skt. dusita.
hvasa jsa bijautta spoiled by vegetation. E 16, 9, bajot-
tanda.
bijsaSa . . . to be poured . . blugs-pa (vaik-, Av.
vaek-, Mid. West Iran. ve^t). *>
binajana . . . to be steeped . . sbans-pa.
binana . . . to be split . . dral-te, Skt. patana. ,
bimji. • ■ sparrow. . . mcbil-pa, Skt. cataka. j
barriysata . . . moved . . . g-yos-so.
bisi, besi . . . buttermilk . . dar-ba.
bisai . . . • expressing locative case in adjectival
form passim. Cf. blvi similarly expressing genitive case,
viysarnjva bisa utca water in the ponds ; purana bisarn acham of
diseases in the womb. So translate passim in Sacu Document :
1. 17, secu bise karntbe the towns in Secu (probably,
according to a suggestion of V. Minorsk}^ ® ^ siei tsiou
the Chinese name, in modern pronunciation Hsi-chou, of
Yarkboto). So also in Sacu Document 20 secu misti kaintha
Secu the capital.
bisu .... bush, tree (used of sin.
castor and juniper)
bithana . . • dril-ba . . . to be tuisted. together.
bitte . . . is cut off . . chod-ba.
biysanai . . • severe . . .mi bzad-par.
biysarnjana . . bkan . . . fill.
3 Sg. Pret. blyslya seized, 3 pi. biysiyaude.
blysma . . • urine . . . gcin passim.
bra, bravi . . . clear . . . gsal-ba.
bramja . . • birch-tree . . stag-pa.
briha, brraha, loc. sg.
brahana . . • back . . . Skt. prstha.
jsimna briha belly, mkbal sked, Skt. kuksi.
brrijsana . . . to be roasted, fried . brnos-pa.
(Baloci brejag, etc.). See aysbrijs-.
bujanana . . . to he extended . . brta-bar bya-ho.
bujsvara . . • pestle . . . btun-bu. E 5, 65,
bajsvarra.
IRANIAN STUDIES V
121
. . . intestines . . loii-ga. Cf. bunu,
E 23, 149.
bqfie hamddri vya inside the intestines.
buysina adj. . . goat’s . . . rahi (Av. buza-),
bvascye, bvastya, adj.
bvastina . . experience . . ses-pa.
byajafia . . . to be melted . . bin (vi-tak-, Av.
vitayti-).
byamjsa . . . astringent . . bska-ba. Konow Suv.
batamjsa.
byara . . . gourd . . . ga-gon, Skt. erviiruka.
byasana . . . to be opened wide gdan
byasde . . . dissolves . bjig-pa (vi-tak).
byura, byurra . . ten thousand . . (Av. baevar-).
ssa byuia juna 1,000,000 times ; dvarabista bjmrra hasta
ysara 228,000 ; bjiira kula jvina a myriad koti times.
So translate E 6, 87, ksei byurru 60,000.
cainbula . . . disturbed. Tib. zi-zi-por, Skt.
akula. pajsa cambula vijisde sin-tu zi-zi-por gyur-pa, Skt.
drster akulata bhrs'arn. Cf. E, 2, 10, bi^sunye carnbule.
cambva, loc. sg. cabvana, cabufia, thicket (according to Skt. and
Tib. parallel passages).
So translate E 25, 420, carnbuve viri.
chai, pi. cha . . sprout . . . myu-gu.
E 14, 153 ; 15, 79 chei ; E 8, 20, chate.
dabausta, dahosta . virility .
ro-tsaba.
dalai, pi. dala . . shell {of egg), bark {of
tree)
sun-lpags.
darana
to be scraped .
bzar.
dasde
. ripens .
smin-pa (Av. dag-).
daujsa
. lead {metal) .
ra-fie.
dide .
sloth
Skt. alasya.
dilaki
. . . little
cbu-nun.
dinara.
dinaryau . Skt. biranya .
gold
disana
%■ V
. fo be thrown .
Fret. 3 Sg. diste. Skt. praksipta.
bor-la.
dista
. . . ripw
smin-pa. See adista
unripe. So translate E 2, 68 ; 23, 294 (cf. Konow’s Review,
p. 36).
dr|ma
. pomegranate .
bal-po sebu. Hence
122
H. \V. BAILEY
Chinese m. d'uo liam (modern t‘u-lin), quoted by Laufer,
Sino-Iranica, 282. Skt. dadima is perhaps in some way
connected.
dramphame, drrauphame exertion . . Skt. vy^ayamaka, Tib.
brtsal-ba.
drrava . . . langind . . . rgod-pa.
Fern, draca in draca tsumata, Skt. cancalarn. E 2, 101 ;
25, 401 dratai ; E 21, 15, drace.
dvisata, dvissa . . two hundred.
dvyana . . . to he beaten . . rdeg-pa.
estavana . . . firm . . . brtan-pa.
estama . . . strength . . . nams-stobs.
Cf. anesta, Skt. asaraka. So translate E 24, 27, nai ne
ysira estatu yindi his heart cannot be stilled.
eysa . . . millet . . . ci-tse, Skt. nivara.
ganam . . . xvheat . . . gro. ( < *gandama-,
cf. NPers. gandum). ganamai *wheat for him.
ganih^na . . . to be made wet.
Ptc. ganaista . . u'et . . . gser.
Cf. ganista loc. sg., Skt. samsveda in the Vajracchedika.
garkha . . . heavy ; honoured . Ici-ba,
yor-yor-ba passim, garkham hvadam khastam jsa, Skt.
gurvaharam.
garkhiittetu
reverence.
gus-pa.
garsa
. throat
Lkog-ma.
gauda, goda
lean
skem-pa, Skt. krsa.
gausa
millet
khre, Skt. priyaiigu
gaysii
. reed
Skt. nada.
ggeiha
. wood
Skt. kastha.
gichauka .
centre, heart
snifi-po.
gisai .
. name of a plant.
glsai
hiya bfita. gisa bfite. rtsva ku
-sa the grass hum.
So in E
2, 13 ; 25, 100 ggl.sa ; E 24,
172, ggisai.
giska
. *rope (in the simile of snake and rope).
grlha
lijim-pa.
mud, clay
gruska
shin, hush, rind
sun-lpags.
Adj.
griiskinai. Cf. E 2, 13, gruske.
guha
. OX ,
ba-lan.
guha salya
. ox year.
guha me
ironnd .
Skt. ksata.
IRANIAN STUDIES V 123
gumalyanii
to be smeared .
. bskus-te (vi-mard-).
Ptc. ggumiilstu. E 22, 11.
gura
vine
. rgun.
gurutca
a brew of grapes
guraka
. uvula
Icehu chuu-ba.
gurma, gurmii, gaurnia pea
. sran-ma, Skt. kalaya.
guysna- .
deer
. Skt. ruru. E 14, 77,
gguysna (Ai
\ gavasna).
gvachame
digestion
. (vi-pak-).
gvacha
. digesting.
garkha gvacha
heavy to digest.
gvachanaka
. nial'ing digest
hjug-par byed-pa.
gvaysde, 3 Sg. .
cracks .
. hgas-pa (vi-vaz-).
So E 5, 110, gvaysda 3 Sg.
gvehaiskye jsa .
. U'ith spoon
thur-ma.
gviha, adj.
. cow's, passim.
gvihaiyii, gvahaiyi
. pain
. zug-niu.
gvlr-
. sag, report
. Pret. guda.
brru liadai pfdiyai pyatsa guda he related it earlg in the
'morning to the parohita.
2 Sg. fern.
guda, 3 plur. gudada. Pres. 3 Sg. gvida. Imperat.
2 Sg. gvera.
Ptc. pres, gvlrada.
Cf. E 24, 141. gviranda ;
E 23. 330,
gverlndi : E 5. 22,
gvlde. So read in Sacu
Document 55, kyi hva hva gviraci.
gyasta-
. cleaned .
. See gyeh-.
So translate in E 25, 490. cu ne gyastu lya what is not
cleaned.
hugyastu
gvehaha ysa^ihanu.
Skt. susodhayitavyara
to be well cleaned.
gyeh-, jeh-
to cleanse, to heal
. Skt. sodhaya-, cikitsa,
Tib. gso-ba.
gyehana . . to be cledoed.
jehiitc. Pres. 3 rh'(in!i : jatte is healed : 3 pi. jebare.
jihari are healed, jehanic, jiliijmc, jehiync. jaluime. jiliiime,
jihiime. Skt. cikitsii ; Tib. gso-ba healing.
jeliiina to be heated.
hadainjsva . . lasting one dag.
ttye ]si pajsa hadainisya /ro/Zi there jive dags' joarneg.
hahalsandau avsmu . dad-par g^air-bas . believing (to hahalj-),
but Chinese version has jc M S--
124
H. W. BAILEY
hajse . . . insects . . . Skt. ksudra.
hala, adj. . . . half . . . phyed.
hala amga . . half the litnb.
halai, adj. . . . half.
halai kamala . . half the head.
halinai . . . crosswise . . yo-bar
halirai, pi. halira . myrobalan . . a-ru-ra (Skt. harltaki,
NPers. halila, ihlllaj ; Kuchean arirak).
So translate E 14, 102, halirau.
hama, hauma . . raw . . . rjen-pa (Skt. ama,
BalocI hamag).
hamai . . . barley meal . . pag-zan.
hamina ..... phag pbye.
hamara, gen. pi. bama-
ram, loc. pi. hamarva joint . . . ts‘igs.
(ham-ar-). A wrong etymology is given, BSOS. vi, 66.
To this base ham-ar-, the partic. is hamicja.
hambamjsya . . general . . . spyihi.
hambarasce . . filling . . Skt. paripurapa.
hambica jsa . . Skt. samasa. . a putting together
See hambrih-.
hambithe, 3 Sg. . makes retain. . sri-bar byed.
harnbithaka, harnbithaka.
hambrrame . . healing (of a u'ound) gso-ba.
hambranaka . . healing . . sel-to, Skt. ropana.
hanibranana . . to be mixed . . sbyar.
hambrihana . . to be put together . sbyar.
Partic. hambirsta, hambairca, hambica. Pret. 3 pi.
hambistamda.
hambusam . . proper . . . hos-pa.
ahambusana . . improper.
handajanatema . . 1 perfected . . yohs-su smin-par
byas-so.
haindevaka . . ripening . . smin-par byed-pa.
hamdrrarnjana . to be kept . . bsruh (hani-drang-).
hamgarana . . to be pulled out . drahs-te, drah-ba.
hamgrihana . . to be bent . . dgug.
hamgujirnde . . *they fear ; hamgvamdum *they feared ; to
hamggujsa- fear, hjigs-pa.
hamgunana . . to be covered . . g-yog-par byaho
(gund-).
IRANIAN STUDIES V
125
hamida
joined .
See hamara.
hajsaro acc. sg.
sphere .
Skt. gocara. Cf. E 24,
51, hamjsarro.
hamjsulyaka : dai ham-
jsuly^a
producing heat
drod skyed-par byed-
3 Sg. hamjsull.
do.
hamjvame
cheicing
mur zin (NPers. zavl-
dan).
hamphajana
to be wrapped
phur-te.
hamphlsana
to he put together
sbyar.
hamthrajana
to be pressed
btsor-ba.
hamthrrista
suffers .
Sam-thag-ba.
hamtranare
they tnake fall out
(used of the hair)
hbj-i-bar byed-do.
haraysde .
extends to
h jug-par hgyur-ro.
harga
easy
bde-bar.
harlysame
sdahs-pa
angry.
haryasa, hirasa, biryasa
black
nag-po.
haryasaurga
blackish
Skt. krsnabba.
haskalaka
a cloth .
ras
haspareksa <st> a
sixty-eight.
hastusa
eighteenth.
hasu, hasa
a swelling
skran-ba.
hasvimda they swell ; hasvame swelliny ; hasvaca, pres.
ptc., stvelUng (Skt. svay-, Av. spay-,
Oss. rasujTin).
hatcyada .
crippled
grum ziii.
hatcyave .
it breaks
cbag.
hauga, hoga
soft
bjam-po. Opposite
to striha.
hausa, hosa
finger .
sor. E 14, 71,
haussu.
hausde
it dries up
bskams-nas.
hauvana
strong
stobs che-ba.
hauvimda
are ground to frag-
kbram-kbrom-du
ments.
btags-ba.
hinai, hainai, henai ; pi.
red
dmar, Skt. aruna. E
hina, hena ; fem.
hatanai, hatigya.
hemja, hija, heje,
hija, himja.
126
H. W. BAILEY
hirafia
condition
sturii liirane, Skt.
sthaulya.
hirtha, hirtha, hairtlia
sudden .
glo-biir.
hisam, liisa
iron
Icags (*asiiaiiya-).
Adj. hlsanai, hi^mje, hlsanlinje.
lilsinaka jsa
with the iron .
Icags-kyis.
hlsinie : gvlhye Insime,
cow's tent
ball! mi-sor.
hlysaniam .
coriander
ho.su, Skt. dhanyaka.
hujlnim, liujija, adj. .
with blood.
huinjuda, adj. .
v'ith blood
klirag.
liiira ; loc. pi. huravua,
thigh
Iji'Ia. So E 23, 168,
huravva.
hura.
liTailiu;ia .
*Cighitrs.
kvainana .
to be dried
skam.s-pa.
hvara
sweet
uu'iar. Konow Suv.
hvarra.
hvarame
eating .
Skt. .sevana.
hvari
*beconte dry .
ca ainga hvari ivhen
ihe limbs become
dry.
hvasix
vegetation
Iduin-bu. E 2, 14,
hvassix.
hvl
U'onKtn .
bud-nied.
iharstii, charsta, aharsta
regularly
rgnin-tu.
inakai, inakii
adjacent
fien-b.skor.
irhva
citron
Skt. niatuluhga, Tib.
klia-liin.
Till
the plant karavirai.
karavirai, si
iru Iliya kata Jraravlra. that is. the root
of ml.
is.sakil
returning
Skt. pratiaivaitaka.
jaste . . ■
part of eye.
liarvasyo jaste Idindil
on the black: part
Skt. krsi.ia-bliage.
sTvo jaste kimdii
on the wkiite part
Skt. sukla-bhage.
jista
boiled
See jsilh-. (Av. yaes-.
etc.).
jsanldii
they boil
skol.
jsanijme
boiling.
jsanafui
to be boiled.
jsiliaia. jsalura. jsahera.
stomach , inside
pho-ba, klion-na. kho-
jselicra.
ra.
IRANIAN STUDIES V
127
kaba . . . « measure of quanlit)j.
dva kaba gviha rrum two kaba of butter.
kahai . . . itck . . . g-yah. Skt. kandu.
kahaitte, kihaitte, 3 pi. itch . . . g-yah-ba.
kj-ibare, kyabari.
kakii, kekii . . powder, oiutment . pbye-iiia, Ide-gu, (cf.
Skt. kalka).
karnii, kania . . wound . . . rma.
kamga, kamgo, karngj-i, skin . . . siiii-lpags, pags-pa. E
kamge. 2, 12. kaingo.
kanipilai, kampile, ka- name of a tree . Skt. kaiuplh'a (kain-
pTlye, kainpilye. pilla, kampillaka,
etc.).
kapuri . . . camphor . . Skt. karpura (Sogd.
kp'ivr, West Mid. Iran, k'pwr, XPers. laifur. Tib. ga-bur).
kara . . . the outside . . pbyi-rol. fien-skor.
kara tcinianani. inig-gi pbyi-rol the part hji}i(j around the
eyes (connection with Av. caxra-, as proposed by Konow,
is unlikely).
karvina, karvina. karvliuitn adj. fien-skor.
karsa . . . (mayic) circle . Skt. inai.K.lalaka.
kas- . . . .be obstructed, fail.
3 PI. dyakyai kasiirc u gva sight and ears fail niig Idons-siii
bon-par bgyur-ba.
3 Sg. kaste, kasti. E 24. 99. kassiru, 2 pi.
kasama obstruction ligag-pa.
kasa, kase jsa, loc. sg. decoction . . khu-ba (cf. Skt.
kasana. kasiiya decoction).
kayacii . . .at night . . . Skt. nisitbe. dinante.
kbaiyai, khaiy^e, kheya illness . . . na-lia (cf. i\Iid. Pers.
jsa. til).
kbaiyTidil . . painful . . zug-ciii.
kbajc . . . mud . . . rdzab, hdain.
E 9, 90 (loc. .sg.) kliarja, E 20, 53, kbarggu.
kbarina adj. . . gla-ba . . . musk-deer.
khase . . . spice . . . spod.
khauy's- . . . move, change . g-yo-ba.
khauysama, kbauvsaunia verbal noun, khoy'sainda pres,
ptc. kliauysde 3 sg. bgynir-bar. akhaustii unmoved, Skt.
acala. khausaiia to be agitated bsgul zin. Cf. askhauysa,
128
H. W. BAILEY —
ttuliostandi (E 23, 308). (Possibly related to Armen, (if
Iranian loanword) seek).
khava . . ■ fomn . . . sbu-ba (Munjani yaf,
Skt. kapba, NPers. kaf). E p. 351, line 9, khava, E 7, 53,
samu kho khava o kho marica.
khaysanya, khaysana intestines . . loh-ka.
loc. sg.
khaysma . . . abscess . . . phol-mig, hbrum-ba. ^
khinda, khendl, khainde like . . . Ita-bu, bzin-du, hdra-
ba, hdra-ho.
khiysara . . . filament . . ze-ba (from Skt.
kesara).
khuna, khijna, khijne, hole . . . sbubs, khuh-bu, bu-
gen. pi. khiinam, ga.
khpnyam.
khusi ... rot ... rul-ba.
Idssana, kissangye, Ids- filled . . . Skt. akirna.
emja, kiseja, kisina.
koraka, kauraka. . wood-pigeon . . thi-ba, Skt. kapotaka.
kralo . • • tin . . . Skt. trapus.
krrirnga-ruvai . . anus . . . Skt. guda, Tib. gzah.
krrimgine aha . . fowl's eggs . . khyim-bahi sgo-ha.
ksarma . . . shame . . . skyehs-pa (in E 25,
495).
ksarma hamya he teas ashatned. E 2, 126, ksarmane.
(Av. fsarama, Sogd. s^'r, s^'rm'k, Oss. afsarm, West Mid.
Iran., NPers. sarm.)
ksausta, ksosta . . serum . . . chu-ser. ,
ysu ksosta pus and serum rnag dan chu-ser. E 9, 33, ‘ ^
nom. sg. ksusta, E 9, 32, acc. sg. ksustu.
ksirinau . . . ninety-six.
ksitii . . . exhausted . . Skt. ksayatam \Ta-
jete.
kuham than jsa . with a cloth . . ras, ras-ma.
kumba . . . flax . . . zar-ma, Skt. atasl.
kuinji . . . yeast . . . rtsabs.
kurkum, kurkum, saffron . . . gur-kum.
kurkarn.
(Sogd. kwrkwnph, Kuchean kurkam. Mid. West Iran,
kvvrksvm, NPers. kurkum.) Jt
IRANIAN STUDIES V
129
kusemate
search
yons-su btsal-ba.
In E passim. E 6, 10, 12, 3 Sg. Pres, kusate, E 6, 9, 112,
3 Sg. pret. kiiysde. 3 Plur. kuysdauda.
kutana . . . to he ground . . btags (cf. Skt.
kuthth-).
kvaysa, loc. sg. kvasa, side of body . . rtsib-log.
loc. pi. kvaysva,
kveysva, kviysva.
dvya kveysva vl in both sides.
-laka, suffix in bakalaka, naukalaka, sikalaka, sikalaka.
laksa .. lac ... rgya-skyegs.
llha .... Skt. leha . . sman-gyi Ide-gu.
Ilka, loc. sg. llkana, in adjectival function after participles, passim.
jista lika kasa boiled decoction.
lodrrai, lodrre, ladrrai . hunter . . . Skt. lubdaka.
loksa, loksa . . rough . . . rtsub-pa.
macamga . . .a measure of quantity zo, Skt. karsa.
mahairsina . . buffalo's . . make (from Skt. ma-
hisa, cf. Toebarian A mahirsan buffaloes).
makala . . . monkey.
makala salya monkey year ; makala re king of the monkeys.
(Sogd. nikr’, Krorain Kharosthi makad'a, Skt. markata.)
maksama
maksana
mamga
mamgara
masta
smearing
to be smeared.
bsku-ba.
pea . . . mon-sran.
. . lasting a long time . rnin rin-du.
1. intoxicated . Skt. matta.
2. besi masta, dar-ma butternnilk.
amasta nye, zo ma-lans-pa. (Mid. West Iran., NPers.
mast, mast.)
(mada-).
intoxicant
Skt. sura
juave . - • intoxication . . Skt. mada (mad-),
mijse . . . woman . . . bud-med.
mijse vi tsQma coition nal-po ; mijsi mirare his wives die.
mina, adj. . . • ff sheep . . lug-gi-
marahya, mirahe . pearls . • ■ Skt. mukta-.
E23, 254, mrahe.
muda sarnga, muda samga, spans-ma . . blue vitriol.
Cf. NPers. murdasang litharge.
VOL. VTll. PART 1.
9
130
H. W. BAILEY —
murana . . . to he rubbed . mnes te, brdzis-pa.
murasa . . . peacock . . . rma-bya.
muyi . . . tiger {in the twelve-
year cycle).
nala viram, nala viram. fistula . . . Skt. nadl-vrana.
namva, namve, nimva, nemve salt lan-ts’va ; adj. namvije (Mid.
West Iran., NPers. namak, etc.),
nara, gen. sg. naira . waman . . . (Av. nairi.)
nasapa, naisapa . . porridge . . ho-thug.
naspastame . . bringing out , . bbyin-pa.
3 Sg. naspaste.
nasphaMaka . . making come out . bbyun-ba byed-do.
naspbusta . . thrown . . . Skt. ksipta.
naspbusta Ida, pbyur-la.
nastausai . . dryness . . skem-pa, Skt. sosa.
Cf. E 6, 28 nastosate.
nastvana . . . to be poured out . dbo-ba.
nasa . . . portion . . . cba.
natcipbak§, natcipbaki subduing . . bjoms-par byed.
3 Sg. patcipbe, nitcipba.
natu . . . deep . . . Skt. garnbbira.
natalsto tsindi, Skt. hesta-gaml.
nauba . . . point . . . rtse.
nauka . . . fine . . . zib-tu.
naukalaka . . . soft . . . hjam-po.
nausa . . . nineteenth.
neha, gen. sg. . . navel . . . Ite-bahi pbyogs.
neba ttausaraa . heating of the navel.
nesta padimana . to be removed . . med-par byabo.
nibejaka, nibijaki . removing . , sel-to.
3 Sg. nibeje, 3 plur. nehejida to ni-halj-.
nibujsamato acc. sg. Skt. nirvana. Partic. nibuta entered nirvana.
E. 15, 10, nilbuta, E. 15, 27, nahuta, E. 15, 27, nibujsanda.
nirausta . . . burst out . . brdol-pa, Skt. bhinna.
nirausta 3 Sg. brdol-pa.
nirujana . . to he burst . . rtol-te.
nirujakyo . . bursting.
nerau . . .a bursting out . brnags-nas rdol-ba.
nirasamdai pres. part, inchoat. . . . rdol-ba.
niskeca, naiskaica . section . . bye-brag (= Skt.
IRA^riAN STUDIES V
131
visesa). Cf. niskici in the Sacu document. Similarly piskici
section bye-brag.
nispasdana . . to he broitght out . byuh-ba.
Cf. nasphaMaka.
naustara jsa . . cutting instrument . mts‘on.
nvasta . . . easy . . . sla-ba, bde-bar.
See anvasta.
nvathanana . . to be cleansed . . sbyans-la.
Cf. E 13, 6, nvanthare. E 25, 423, nvamthare. E 25, 117,
nuvamthata. E 21, 14, nuvamthanda. E 21, 14 ; 24, 30 nvamtte.
nye, ni . . . curds . . . zo, Skt. dadhi.
nyetutca, hetutca, nyevutca, nivutca, zo kha chu, a brew of curds.
nasa . . . reduced, humble.
cu buijasa nasa sira padimi tclien the voice is reduced, it
restores it.
nasa bisa
. humble servants.
nasika .
. weak
. zan-pa. Cf niyassa-
nyassa-.
nustyana .
. to be ivrapped
. phur-la.
nustana.
nustana. Partic.
nustye, nustai (ni-pasta-
iua > }ii).
nuska
. accustomed
. goms-pa.
anuta
. unaccustomed .
. Cf. nyuta-. (ni
yaug-).
pa
. then
= patca passim.
pachai
to he cooked .
. btsos-te.
pachare
they cook
ts‘os-par byed-do.
Cf. pajs-.
padajsame
. burning
. bsreg-pa.
padasda
. it destroys
. h jig-par byeddo.
E 12, 36, padasda.
padausidi ; cu hiyara padausidi. *u'hen the fruits are ripe.
paha, paha ; zu-ba chyme, khu-ba liquid, kha nas chu spittle.
paherang, . . . to be steeped . . sbah zih.
pahastana . . . removed . ■ Skt. vivarjita.
E 7, 10, 27, pahasta.
• cf. pihisame . . obstruction . . hgags-pa.
paja- . . . special.
pajarn dharmarn. Skt. guna-prabheda the (64) special
properties.
132
H. W. BAILEY —
dharraa paja the (18) exclusive properties. Skt. avenika-
dharma.
pajarrustada, 3 plur. Pret. they surrounded. Skt. parallel text vestita.
3 plur. Pres. : brre garsa nurahe pajarustlda the pearls encircle
the beloved neck.
pajs-
. to cook .
(Av. pak-, etc.).
paste
. is CMoked.
pajsama, verbal
noun, brtso-ba.
pajsaka, adj. partic.
pajsana, partic.
necess. Cf. pachare.
pajsa, pajse, pejsi
. greatly, passim
cher, sin-tu.
pajukiiiai, adj. .
. having a lid .
(gund-, cf. pa-justa).
pajukinai bajam, snod kba.
pajukausta bajam
. a vessel xvith a lid.
pajusta
. finger-ring
Skt. anguli-mudra.
paka-
. *foot. .
. Cf. puka-.
di paka .
. under the feet.
pakva orga
. honouri)ig the feet.
pakarn aurga
. honouring the feet.
pamuhi, pamvaha
. clothing or equipment ( < *pati-mauxsa- ;
pati-mauk-
to put 071, cf. vimuha deliverance). Cf. E 25, 277,
pamatu. E 25, 406, pamatanda.
paniysau .
. urinary organs
. cbu-so.
paraba, pariba
base, foundation
• gzi-
avarabe ivithout base, to parautta
supported.
paramga .
decrease
. bbri. E 24, 109,
paramggu.
parainjsana
to be lessened
pbri-ste.
pararnjsa vastTda
they come back
. log cin. Cf. E 13, 45,
paramjsa.
parcblysde
. it sinks
Skt. vilambisyati.
parcbuta .
destroyed
. Skt. upahata. (Cf.
Mid. West Iran, patkoft). Cf. tbe -cbus
- of E 15, 54, pacbusinda.
E 6, 41, pacbusta.
parehana .
. to be kept
bsrun zin.
parlsame .
decreasing
dbri-ba. Seepararnjs-.
parlysma .
. urine
gem.
parka
top
Skt. sirsa.
parva
. behind .
rtin-pa las.
parivi ana
.
rtin-pa nas.
IRAXIAX STUDIES V
133
parvaste . . .is ripened . . smin cifi. Cf. par-
vacha. (pari-pak-)
parysa, parysyau . servant . . . Skt. gukyaka.
gyastii klya parysa amaca hamfda servants and ministers
of the Icing {god = Skt. deva) together. Cf. E 2, 50. parvsa.
paskosa, paskausa
Skt. anaha
. Ito sbo-ba.
pasojana
to he washed
bsal zin.
pasa halai
evening
. nub-mo.
pasamjsya rva .
autumn
Skt. sarad.
pasi, pasa .
pig
. pbag. Probably also
pyasa.
pasausde .
is obstructed
. bgag cin.
paskyastii
again
•
pastista
produces
. bskj’ed cin.
pastrisame
stiffness
. rens-pa.
pathana
to be burnt.
Pret. 3 pi. pathutandii, bsregs
-so. Partic. patbutu.
pataraho acc. sg. .
object
. gnas, Chinese
E 21, 69, patarahva, Konow Suv. patarahe.
patarotta .
supported
. Skt'. samsrta.
Cf. parautta.
pamtsana .
to be avoided .
. span-bar bya-ho.
paysanu
side
. hgram-logs.
C O
paysata, paysava
*cognate.
pura paysata
*oivn son.
paysavye hvarakp
*oivn sislePs son.
pujai.
perai, pi. pera .
plague .
bla*gnan.
perai avamdu achai.
phaja, loc. sg, phajina
oven
. me-mar-miir.
phanai : pasta phanai
rkan gyi phug innermost part of the palate.
phara, loc. sg. pharana
water-pot
. Skt. ghata.
phisana
to be avoided .
. spah-ba.
phiysgam, loc. sg.
urinary organs
. chu-so.
phiysgafia.
pi ... .
fat
• zag-
pagajsa, adj.
having poicer .
stobs can.
pihisame
obstruction
. hgags-par hgyur-ba.
3 Sg. pihista
it stops.
pila
calamity
. Skt. pida.
134
H. W. BAILEY—
pira, plra, gen. pi. insect . . . srin-bu, hbu.
piranam.
piruda, adj. . . having worms . srin-bu dan.
pirauraka, piroraka . removing . . bsal-ba.
3 Sg. pireda. Pret. E 21, 32, puraude.
piskelyame . . removing . . bsal-ba.
piskalyana . . to be removed.
Pres. 3 pi. paskalinda E 15, 36.
Partic. piskici, peskece, piskista, bye-brag section.
pisalyana . . . to be smeared . bskus-na.
pisalyama, verbal noun.
pitciyi . . . spoon . . . tbur-ma.
prabauyana . . to be put on . Skt. pravrtya, pra-
varitavyani. See Konow Suv. s.v. prahos-.
prara, prrara . . nature . . . ran-bzin.
bina-vrrara . . having nature of wind rlun gi ran-bzin can.
prisama-, presama . suitable . . . bphrod-pa.
pribaram . . . knife . . . gri (Skt. prabarapa).
prruyi . . . a measure of quantity.
dvi dvi prruya . two prruya eadh.
pub!, gen. sg. pubya . purohita.
punvana . . to be put into . brdzans-la.
puram, loc. sg. purana womb . . . mnal.
puritta ... ill ... na-ba daii.
pvata, pvava . . cool . . . bsil-ba.
pverame . . . removal . . . bsal-ba.
pvisana . . . to be covered . . g-yogs-pa.
p vista . . . covered . . . g-yogs-pa.
pviysaka . .a wrap . . pbur-te.
pyada
pyadai jsate . . comes hack.
pyada-tsuka . . turned aside . . log-par son-ba, Skt.
unmargT.
pyada-garyau sanyau *with enemies whose acts are perverse.
pyada-vadi . . *walking on perverse course.
Cf. E 18, 19, pyada : kho ju masta bastii pyada.
pyabana . . . to be broken . . bcad-pa. Partic. pya-
basta, pibasta.
pyama, pyama . . covering . . g-yogs. (to pa-tam-).
rraga . . . *side, hank.
IRANIAN STUDIES V
135
ttaja hye raga bidi . on the river's bank.
bavana raga vi basa. in the garden beside the palace.
raba .... pain . . . na-ba.
rrai E 2, 13 . . *plain . . . (Sogd. ray Trebiov,
Slid. West Iran., NPers.ray»«eadoMi).
E 7, 52 ; 10, 8 rrau ; E 15, 49, rra ; adj. E 20, 7 rracye.
E 10, 8, o ttarra rrau vata aska. or thirsty deer on the plain.
rajsina . . . decrease.
rajsina padime, hbri-bar byed-do.
rrapbai jsa . . fighting . . Skt. yuddbe, Tib.
stobs hgyed-pa.
rrasana . . . to be straightened . sran zin.
rrasa . . . green-grey . . sno-skya, Skt. syava.
rrima, loc. sg. rrimana faces . . . dri-ma.
rrlpe . . . chief queen . . Skt. agra-mahisl.
risa . . . appetite . . yi-ga.
rlyai . . . anus . . . rkub.
rraysduii'a. Skt. rajaduhitar-. Cf. Konow Suv. rraysduiranu.
rriysi . . . ^trembling.
yslrg, rrlysj . . trembling of heart.
rriysutca . . . brew of rice . . hbras bskus-pa.
rrlysva, adj. . . of rice . . bbras kjd.
rujai . . . belching . . sgregs-pa.
rausta . . . breaks out . . idol cin.
See niruj-.
rruna, rrunu, rrurp . oil, fat, passim.
gviba rruna . . butter . . . mar. E 2, 16, rruna.
In Samgbata Sutra 14a, 4, read gvibu *rrunu o kujsatinau butter
or sesame oil.
kumjsatlnai rrunu . sesame oil . . bbru mar.
(Av. raoyna ; Sogd. yaw-royn. Mid. West Iran, royn.)
rrijnai, rrgnai . . madder . . btsod (Mid. West Iran.
roSin, BalocI rodan, NPers. ruyan, runas).
rrusa, rrusa gen. sg. . barley . . . nas. E 14, 91, rrusa,
rruse, rrusi 14, 90, 93 rruso.
rrusada . . barley flour . . nas rjen-par btags-
pabi pbye-ma. See ada flour.
rrustira . . . the plant arka . Tib. arka.
adj. rrustarine.
rrustiram biye ysice pera yellow leaves of the arka.
136
H. W. BAILEY —
rrutam jsa
wan-, BalocI ro0,
^a . . .
adj. ^vine, sa
imtestines
Mid. West Iran, rwtyk,
copper .
,vina, fem. savimje.
^lai
^lanasta
^andramata
samga
sasvam
adj. sasvaninai.
solakya, ^alakyi .
saulana .
siyi, siye
slyanrga
sikara
s^alisma
^imusai
simga . ' .
ssiphisa
sQdasi
suhima
suhyame
sQhyana
iva ....
man u sva utca
saukala
seraamgye
sikalaka
sisda
Ptc. sista (srais-).
sasgye yadanda
E 24, 28, ssasje.
skala
gvana skala .
stuka
su .
adj. svinai.
sunana
sipta).
the one side
to the one side .
name of goddess
shell
mustard
absorbent
to be absorbed.
white
whitish.
sugar
phlegm .
spoon
half a prastha.
yellotv .
eleven.
I make .
verbal noun.
to be made
half
wine and half water.
rheum .
slimy
young
takes hold
they did homage
* blow .
* a blow on the ears,
being
horn
to be put in .
rgyu-ma. (Av. uru0-
NPers. ruda.)
zans.
(sa-(li)alai).
pbyogs gcig-tu.
Skt. sri.
dun.
yuns.
bjib-pa.
dkar-la, skya-ba.
kha-ra.
Skt. slesma-.
thur-ma.
Skt. pandura.
Skt. prayojayami.
byas-de.
phyed.
rnu-ma.
Skt. piccha.
Skt. bala.
bdo-bar byed extends.
rim gro bskyed-nas.
Cf. E 24, 160, skalu.
gnas-pa.
rva (Av. sru-, etc.),
smyugs ( = Skt. nik-
IRANIAN STUDIES V
137
sijDa, loc. pi. sijiiva . loins . . . Skt. katl (Av. sraoni,
etc.). E 21, 45, ssuSi.
sura . . . saline . . . Ian-ts‘va.
(Mid. West Iran, swr, NPers. sur.)
svaka, svakye . . pastil . . . ren-bu.
svida . . , milk . . . ho-ma (Av. ;:^svipta,
Mid. Partb. syft, Zaza sit.)
adj. svidinai, svidausta. Cf. Bacu Doc. 57.
syanan^ • • . to be poured . . blugs-na.
®uda .... cold . . . gran-ba (Av. sarata).
samanaka . . correct . . . ran-par.
sambajatu . . may he prosper . 8kt. samrdliyatu.
3 Plur. sambajandu. Skt. samrdhyantu.
3 Sg. opt. sabajiya.
sameyaSa . . . to be mixed . . sbyar-ro.
samida . . . they agree . . mthun-te. E 7, 16,
samindi. E 13, 123, samate.
Sana, sana . . a medicinal herb . la-la-phud.Skt. vavani.
sani, s^nii . . . excrements . . pb}-i-sa.
sanaida . . . luxuriant . . snuin-pa.
tcamjsai sanaida . his hair is luxuriant.
sarndvaina jsa . . in combination . bdus-pa (corresponds
to the Skt. medical term sannipata). Cf. E 23, 217, sarndavata.
samga . . . stone . . . rdebu.
sainskalanu gen. pi. . dharmus . . chos, Chinese j^.
soyana, sauyana . to be rubbed, ground bdar-te.
sibista . . . nimblc-tongued . Ice bde-ba.
sijascye . . . accomplishment . hgrub-pa. Skt. siddhi.
sera, saira . . a measure of quantity srah. Skt. pala.
siravatinaina suhana . leith love . . Skt. priti-.
sairkha . . . retention . . sri-ba. bsdus-pa.
sivame . . . enjoying . . zos-nas. Skt. sevita.
siyanu rre . . king of siya-birds . hah-pabi rgyal - po
® I (Skt. hamsa-raja.) A'so syanii rre.
skara . . . coals . . . Skt. ahgiira (Pasto
skor, XPers. sikar.)
tauda skara . . me mdag. . . glowing embers
spaju . . . alum . . . klia ru ts‘a.
strajanaka . . retaining . . sri. hgag.
strajaka 3 Sg. straji mi hb\'uh-ba.
138
H. W. BAILEY —
striha, streha, straha . hard, vehement, excessive passim, drag-po,
bdo-ba, to strang- in straj- and stris-, pastris-. (A wrong
etymology is proposed, BSOS. vi, 60, note 1.) E 23, 146 straha
dumei his tail is stiff (Cf. Munjani troj : trayd- to fasten,
NPers. taranjidan to compress.)
strisame . . . stiffening . . rehs-pa.
strisanana, causative.
strisida, strisidi, 3 plur. .... rehs-pa. Inchoative
to strang-.
suli, pi. sulya . . *man of Suli (Tibet. Su-hg = Kasyar).
surai, sura, aec. sg. surau, instr. sg. suraina, inst. pi. surv'o, loc. sg.
surana, suna, plur. sura, fern. acc. sg. surgyo (and surjo Konow
Saka Studies, s.v.) clean, Skt. suci, Tib. gtsah, dag-pa. Cf.
asurai not clean. E 23, 126 ; 25, 219. (Av. suyra- red, Skt.
sukra- white.)
suraka, loc. sg. surakana clean . . . gtsah-ma.
sutta . . . acid . . . ts'va, Skt. sukta.
svana girai . red ochre . . btsag.
ttaga. . . . tree . . . sin.
ttagutta, ttaguttau, ttagutto, ttaguttarn, ttaguttam, ttaguttyau jsa
*of such, of the same.
pane hvamdye ttaguttam haudi 4emgam for each man, of
these, seven simga in quantity.
ttahirai . . . correct . . . yah-dag (= Skt.
samyak).
ttahira . . . excellent . . des-par.
ttaja ■ . . . * river.
nva ttaje mista katha gidagitti nami along the river is a
great city called Gilgit (the identification of Gidagitti is due to
Professor F. W. Thomas).
ttamga . thin . . srab-pa (Mid. West
Iran, NPers. tang narroic, tanuk thin).
ttanvame . . . tantra . . . rgyud (Av. tan-, etc.),
ttara . . . partridge . . sreg-pa. (Cf. Sogd.
ttr’w, NPers. taSarv, “ Median ” rirapos).
ttasta 3 Sg. . . flows . . . hdzag-pa (Av. tak).
ttasta u byasde . foivs and dissolves.
ttaura, ttora, tturai 'nwuth . . kha.
ttausama . . heat . . . ts‘a-ba.
ttausace, pres. ptc. E 24, 123, ttausai.
IRANIAN STUDIES V
139
ttavamdya . bile . . . mkhris-pa, Skt. pitta,
ttavaindyuda . . arising from bile . mMiris-palii.
ttavare, ttevare . are hot . . . ts‘a-ba.
ttavai . . . fever . . . rims Skt. jvara.
tca^a . . . *beside.
suka vara riaista sudana tcada viysajai alone there sat
Sudhana beside the pool.
So translate Sacu Document 62 tcadi mahasamamdrri beside
the ocean.
tcaulasa . . . fourteenth.
tcohoruvaretcoholsuvo, loc. pL, forty-four.
tcalca-, makasamudra tcalcanu, Skt. samudra-paryantam. Cf.
E 7, 7, tcalco. E 7, 9, 32, 43, tcalca.
tcarnjsa . . . hair . . . skra.
E 23, 146, tcamjsi kada mulysga his hair very short.
tcata*kva loc. . . a ivell . . . Iten-ka, Skt. tadaga
(Av. cat, Sogd. cat).
tcijsa, loc. pi. tcijsva . breasts . . . nu-ma.
tcrjsva, svlda . . milk in the breasts . nu-zo.
thamjlna . . . to be brought out . pliyuk-ba.
thyauta . . . formerly . . sna-ma.
ttira . . . sour, bitter . . skjiir-ba, kha-ba.
(The si.N tastes are : dajsamdai, ttira, byamjsa, ahvarai, sura,
hvara). Konow Suv. ttira, Skt. tiktah bitter. (Cf. ilid. West.
Iran, thl *ta;^l, NPers. taly, Pasto triy ; Skt. takra buttermilk,
details in Jlorgenstierne, Etymological Vocabulary of Pashto,
p. 83.)
trihe, ttrehe, ttraharn radish . . . la-phug.
ttriksa, ttraksa, treksam harsh, sharp . . rno-ba.
ttrola . . turbith {ipomoea turpethum) Skt. trivrt.
(ttrola? < *travula, cf. Skt. triputa, Tib. dur-byid, NPers.
turbid).
truysa . . . gourd . . Skt. trapusa.
tslsl . . .a title, cf. Tib. ts‘i-si, JRAS. 1927, 817.
ttumgare . . ginger . . . bcah-sga, Kuchean
twahkaro.
tvada, tvadare . . more . . . Ihag.
tvane . , . produces . . skye-bar byed cih.
tvanaka, verbal adj.
ula, aula, ula . . camel . . . (Cf. Krorain
Kharosthl uta.)
140 H. W. BAILEY —
adj. ulina
camel’s .
rha-mohi, Skt. austra.
uskatta
above
shar.
cu uskatta hva
as stated above.
ustamjana
to be uttered .
Skt. uccarayitavya-.
uvi
mind
bio.
uysna
breath .
dbugs.
uysanam plur.
nature .
ran-bzin.
uysdimana
to be cooled
bsgrahs-pa.
Cf. aysdem- (dam- to bloir).
vaha
pastry .
mes btsos-pa.
rrijua vaha, snum khur pastry cooked unth suet.
vahajafia .
to be given
dbul-lo.
vahasta
descended.
aysanyau vahasta. Skt. utthaya asanebhyah.
vahaiysana utca
falling water .
bab-pahi chu.
vameys^na
to be hoisted together
dril-ba.
vasarum, vaseru
gout
dreg, Skt. vatasonita.
vasana
tnethod .
thabs.
vaskalyfma
moment .
Skt. nimesa.
vatcaste
sprinkled
Skt. sikta. So in E
23, 140, vatcisde
he sprinkles.
vathamjana
to be left over
lus-pa, hdug-pa.
Pret. vathlye.
vathanara .
nurse
nad g-yog.
vathavana
to be placed .
Skt. utthapya.
vausanidai
swooning
myos-pa.
vausai
swooning
myos-pa, Skt.
murccha .
vihilai
terminalia bellerica .
skyu-ru-ra. (Skt. vib-
hltaki, vabheda.
Chinese lit ^ ^
modern pronunciation
p‘i-li-lo, XPers. balila.)
vijsyame
seeing
mthon.
3 Sg. Aujsyate.
vilaka, valakam
little
chuh.
So translate vilaki, velaka. See Konow Saka Studies. The
etjTiiology, BSOS. vi, 61, is wrong.
vina, \dna
illness .
na-ba. Cf. vlyane.
Saka Studies.
vinausta, vinosta
ill
na-ba, nam-thag-pa.
vdtkavlje, adj.
of children
Skt. kumara.
IRANIAN STUDIES V
141
viysamjva loc. pi.
ponds.
viysamjva bisa utca
water of ponds
rdzin-buhi cbu, Skt.
vapya.
viysama
improper
mi bpbrod-pa.
vrri
emema .
Skt. vireka.
vyach-
retnove.
vyachidi harabisa pile *They get rid of all calamities.
pila harbasa vyacbai, Skt. parallel text vidhuta-papa.
khva jsa vyachiinda ksira acha pile.
yarnda
continually
rtag-tu, Skt. muhuh.
ysaiysa
bile
mkhris-pa.
ysalva
curcuma
skyer-pa.
ysanu, ysanve .
knee
pus-mo.
ysarabasta, ysambaste
white garlic .
sgog skya.
ysara
partridge
sreg-pa, Skt.
kapinjala.
ysautta,
flows
zags-pa, bdzag-pa.
Cf. E 4, 142, ysotta. E 25, 206, ysautta.
ysidai
yellow .
Skt. pap(Ju, Tib.
ser-ba.
ysidaurga
yellotvish. Fem. ysica, ysicaurga (Av. zarita).
ysinaban^
to be washed .
kbrus bya-ba.
ysira, yserai
ochre
Idon-ros.
ysira, ysirra
rough
rtsub-pa.
ysirai, ysairai
arsenic .
ba-bla, Skt. sila.
ysita aysmuva
zum-pa.
dread
ysu, ysu .
pus
rnag.
ysu u ysica utca. Cf. ysu ksosta. E 9,
32, ysu.
ysuma, ysqmi .
broth
^-kbu.
ysunaka, ysubaka.
cap
ts'ags.
ysuna, ysve.
ysiinana
to be strained
btsags.
ysunamate
asrava .
zag-pa.
ysijnamdai
flowing .
Skt. -sravi.
ysvye
taste
ro.
loc. pi. ysvyanuvo.
142
IRANIAN STUDIES V
Additional Note
The interest of the words saira (sera, siri) and ssandrdmata is
sufficient to justify two brief notes.
(1) saira translating Skt. jmla (Tib. srah) is for older Saka *saterd
with the usual loss of intervocalic -t- (cf. sate, se, ssa “ 100 ’"). This is,
of course, the well-known word attested in Arm. sater. Mid. West Iran.
styr (to be read sater, JRAS., 1930, 17), Sogd. styr, NPers. 'styr,
Arab.-Pers. 'sfr.
In meaning the nearest is the sadera of Krorain Kharosthi, in the
prescription No. 702 ; —
susm-ela dhane 1 sakara sadera 4
‘ one ddnaka of sukstmild, four sadera of sugar ’.
(2) ssandrdnmta < *sanddr>mti- < Old Iran. *maritdrmati- (the
metathesis similar to Saka grdma ‘ hot ’, Old Iran, garma) corresponds
to Avestan spmtd drmaitis, and attests the word in the expected
Saka form with ss- < su-, beside the word smndai < Old Iran.
suantaka-, discussed in BSOS., vii, 288 seq. South-west. Iran. (Pers.)
*santdrmati and this form in Saka show that the word existed outside
the Zoroastrian (Avestan) texts where sp < sii. Hence either (1)
the phrase and meaning (for which the most satisfactory original
meaning is ‘ mighty foundation ’, the drmati- being attested in Arm.
armvxt ‘ basis ') are pre-Zoroastrian, and of the Iranian period before
su became sp- in some dialects (a period to which the suggested meaning
would suitably belong), or (2) the dialectal form of the Zoroastrian
texts with sp- has been received as a compound from Zoroastrians,
but replaced by the appropriate dialectal form, implying that the
speakers knew the regular corresponding forms in their dialects with
ss and s respectively. If the first alternative be preferred, *suantd
drmatis will be a designation of the earth in pre-Zoroastrian times.
Then in the Gathas this word has been by the usual method of popular
etymology brought into connection with another word *aramati-
‘ right thinking ’ unattested in Iranian texts (since the attempts to
find it in Av. drmaitis are unconvincing), but in Vedic ardmati. The
verbal form is found in Yasna 45, 11, arim mainydtd, and this meaning
in the later commentaries in bavandak-menisnih ‘ perfect thinking ’.
The old, original, meaning of *suantd drmatis (probably ‘ mighty
foundation ’) then appears in Avestan texts outside the Gathas,
where spantd drmaitis is used in invoking the earth.
A History of the Caliphate in the Hayat al-hayawan
of ad-Damiri
By Joseph de Somogyi
I
O WING to the voluminousness of the great Arabic works on
general history, from the fourth century a.h. onwards many
abridgments were made from them by both the authors of the large
works themselves and later hands. These concise works * contain a
brief survey of the history of Islam as is seen in the KitCib dimal al-isldtn
of adh-Dhahabi or an enumeration of the illustrious persons of a
longer period as is seen in the Tabaqal al-kuffdz of the same author.
The great popularity of these compendious works can best be seen
from the fact that in some cases, such as the Muriij adh-dhahab wa
ma‘ddin al-jawaMr of al-Mas‘udI and the two works of adh-DhahabI
mentioned above, they were known earlier in the Occident and are
still more used in the Orient than the voluminous original works
from which they were abridged. No doubt the necessity for such works
accounts for the inclusion of a history of the Caliphate in one of the
most remarkable encyclopaedic works in Arabic literature, the Hayat
al-hayawdn of the Egyptian Abul-Baqa Kamaladdin Muhammad ibn
Musa ad-Damirl (from 750/1349-50 or 742/1341-2 to 808/1405-6).^
Even those who are accustomed to the compilative character of Arabic
literature are likely to be astonished at the fact that such a work, a
zoological lexicon as its title implies, contains a survey of the history
of the Caliphate. Ad-Damlrl's work, however, is not to be considered
as a zoological reference-work in the modern sense of the word, for,
treating as it does, in alphabetical order of all the animals known to
the Arabs, it also includes plenty of traditions, Qur'anic citations,
and proverbs as well as excerpts from the works of Arabic historians,
theologians, scientists, physicians, and interpreters of dreams, that is
to say materials which are mostly connected very loosely with the
1 For details see my paper " Ein arabisches Kompendium der Weltgeschichte.
Das Kitab duwal al-islnm des ad-I)ahabj,*’ Islaniica, vol. v, pp. 334-353.
^ See my “Index des sources de la Ilaydt al-hayawdn de ad-Damiri”, JA.,
juillet-septembre, 1928, pp. 5-12; and A. S. G. Jayakar, “ Ad-Damiri's Hayat al-
Hayawdn {A Zoological Lexicon), Translated from the Arabic,” vol. i. Introduction,
pp. i-xxx, London-Bombay, 1906.
144
JOSEPH DE SOMOGYI
animals in question : in many cases they are quoted simply because
the name of a certain animal once occurs in them. Thus the Hay at
al-Jiayau'an is a profuse and well-nigh inexhaustible treasury of Arabic
folk-lore, the more so as it frequently contains references to authors
or works who or which are not known to us from other sources that
are in general use.^
Small wonder, therefore, that ad-Damiri‘s work also contains
shorter or longer digressions on topics which are quite out of place in
a zoological lexicon, among them a History of the Caliphate, which
makes one-thirteenth of the whole Haydt al-hayawdn? But even
those who are accustomed to the many-sidedness of ad-Damlri will
probably note with astonishment that he relates the history of the
Caliphate under the heading al-iivazz (the goose), and will ask why it is
not treated of in the article al-insdn (man) as would be more reason-
able and plausible.
In the article al-iwazz ad-Damlrl, after giving a short biography of
the poet Abu Nuwas and some short remarks on geese, relates,®
after the Mandqib of the Imam Ahmad who had it on the authority of
al-Husayn ibn Kathir and this latter on that of his father, that one
day when ‘Ali came out for the morning prayer the geese began to
quack in his face. The people round him tried to drive them away,
but ‘All said : “ Let them alone, because they are wailing ! ” The
Kharijite Ibn Hidjam then stabbed him, and after al-Husayn ibn
Kathir’s suggestion that they should settle with the Yamanite tribe
of Murad, he answered : “ No, imprison the man ; if I die, kill him ;
and if I live, there is ‘ for wounds retahation * Then, with reference
to Ibn KhalHkan’s Biographical Dictionary (Wafaydt al-a‘ydn), ad-
Damlri narrates the reasons and the circumstances of the act of
Ibn Muljam and briefly quotes the different opinions about the situation
of ‘All's grave,® of which subject he treats in more detail, and with
reference to Ibn Khallikan, also in the article al-fahd.^ Then there
follows a separate section under the title Fd'ida ajnabiyya (foreign,
i.e. remarkable, information), in which ad-Damiri states in four lines
of rhymed prose that “ tradition has ramifications (i.e. is susceptible
^ For the sources of the Hayat aUh/tyaunn see my “ Index ”, pp. 13-128.
2 In the third edition of the Hayat al-hnyamin al-kuhra^ Cairo, 130o a.h., in vol. i,
pp. 44-96.
® See Hayat al-hayawdn^ vol. i, p. 43, 11. 22-5.
* al-Qur'dn, v, 49.
® See Haydt ahhayawdn , vol. i, p. 43, 1. 26, to p. 44, 1. 19.
* Ibid., vol, ii, p. 196, 1. 28, to p. 197, I. 9.
^ Ibid., vol. i, p. 44, 11. 20-1.
A HISTORY OF THE CALIPHATE
145
of several interpretations [s/iMjMw]), and the imparting of knowledge
verifies to the seekers what they hope (to find), and renews for them
what the wicked forget considering the hard times ; I desire to
mention here a strange information which historians have related ”,
namely that every sixth person who reigned over the Muhammadans
abdicated or was deposed. To prove this assertion ad-Damiri sets
out to describe what historians have mentioned about the caliphs, to
which he wishes to append a httle “ about the life of every one of them,
his times, the cause of his death, the period of his reign and his hfe
in order to complete that information and for (general) advantage
and good
After this introduction ad-Damiri relates the history of the Cahphate
beginning with the prophet Muhammad and continuing with the
first four caliphs, the Umajyads, the ‘Abbasids, and the Fatimids
of Egypt, the author's country. A separate subchapter is devoted to
every one of the caliphs, the title of which for the sixth one usually
bears the indication iva hmva ’s-sddis ^m kJiuli'a (or wa Jchuli‘a
u’a qutila) Jcama saya'tl (“' and he is the sixth one and was deposed ”,
or ■“ deposed and killed ”, as the case may be, *• as follows ”).
The most characteristic feature of ad-Damiri's History of the
Caliphate is its brevity ; he aims at conciseness in his narrative. In
each chapter ad-Damiri first relates the election of the cahph
in question, then the chief events of his reign, then he describes his
good or bad quahties, with special emphasis on whether he was a good
Muslim or not, and finally he records how and when he died, and gives
the number of years of his fife and his reign, and relates his burial.
Among the events of Islam the rehgious struggles with the
Kharijites and the Mu'tazihtes are especially remembered. Political
events are only briefly described. In writing of the prophet Muhammad
ad-DamIrl is content with enumerating his excellent qualities and
stating that the learned have composed many books regarding His
life, His times, His mission. His campaigns. His qualities. His miracles,
and His good and amiable deeds ; if we desired to describe even a little
of them, it would take several volumes : but that is not our purpose in
this book.” ^ About the military merits of ‘Umar al-Faruq he simply
states that ‘‘ he made vast conquests, took large provinces, and was
the first to be styled ‘ the Commander of the Faithful ’ ”,® and in the
later part of his record he only briefly enumerates his conquests *
'■ Ibid., U. 22-3. * Ibid., p. 45, U. 4-6.
“ Ibid., p. 46, U. 15-6. * Ibid., p. 48, U. 5-12.
VOL. VIII. PART 1.
10
146
JOSEPH HE SOMOGYI
and also those of ‘Uthman ibn ‘Alfan in the subsequent chapter.^
^Yriting on the cahphate of al-Walid ibn ‘Abdahnalib, he briefly
mentions that “ during his caliphate some important conquests were
made, such as those of Sind, India, Spain, and other well-kno^vn
places But all the more plentiful are the illustrative instances of
the good or bad qualities of the caUphs which he always records in
detail. And in order to prove his theory that every sixth imam either
abdicated or was deposed by force and killed. ad-Damiri also expounds
the manner in which the caliphs, especially the sixth ones in the order,
ended their reigns. Beginning with the prophet Muhammad, the sixth
cahph, al-Hasan (40-41/660-*662), abdicated; counting from him
onwards the sixth one was ‘Abdallah ibn Znbayr, the counter-caliph
of al-Hijaz and ‘Iraq (64-73/683-692), who was dethroned and slain,
and the sixth one subsequent to him, al-Walld ibn Yazid ibn
‘Abdalmalik (86-96/705-715), was also deposed. As, however, only
three caliphs of the Banit Uma}ya succeeded him and there would,
therefore, be a breach in the order of every sixtli caliph abdicating
or being deposed,® he had to count the ‘Abbiisids from their
founder onwards. The sixth ‘Abbasid cahph, Muhammad al-Amin
(193—198/808-814), was deposed and slain, the sixth one after him,
Ahmad al-Musta‘In billah (247-256/861-870), abdicated and was slain,
likewise the sixth one after him, Abul-Fadl Ja'far al-Muqtadir billah
(295-320/907-932), who was dethroned twice ; the si.xth after him
was Abul-Fadl al-Muti‘ billah ibn al-Muqtadir (334-363/945-974),
who abdicated, but if Ibn al-Mu‘tazz who reigned only one day is
not counted, the sixth one is his successor, Abu Bakr ‘Abdalkarim
at-Ta‘I ‘billah (363-381/973-992), who also abdicated. After him the
sLxth caliph was Abu Mansur Ja'far ar-Kashid billah (529-530/
1134-6) if Ibn al-Mn‘tazz is not counted, but if he is counted, then
his predecessor, Abu Mansur al-FadI al-Mustarshid billah ibn al-
Mustazhir (511-529/1117-1135), was the sixth one ; both were attacked
and slain by the Assassins. This rule, however, cannot be maintained
in the case of the last ‘Abbasids, because neither al-Mustansir billah
nor his father, az-Zahir bi amr Allah, who followed the two last-
mentioned caliphs as sixth ones, were deposed or abdicated, bnt ad-
Damiri finds a rather ingenious explanation to carry his theory
through : “ . . . during their reign the power of the Tatars increased
‘ Ibid., p. 49, U. 20-2.
= Ibid., p. 61, II. 29-30.
•< Ibid., p. 68, 11. 28-31.
A HISTORY OF THE CALIPHATE
147
and their affairs took a serious turn ; they conquered an extensive
portion of the country of Islam. Jalaladdin Khwarizmshah was lost in
the time of al-Mustansir in a battle which was fought between him and
the Tatars, and this (event) was a greater and more serious one than
dethronement.” ^ After these two caliphs the rule ceases to be
appHcable to them, they having been succeeded only by al-Musta'sim
billah whom the Tatars killed in 656/1258.
After the extinction of the ‘Abliasids there were no caliphs for
three years until 659/ 1260-1, when the Eg}'ptians took the oath of
allegiance to al-Mustansir billah.
Among the rulers of the Eg^'ptian Eatiraid or 'Ubaydi dynasty
ad-Damiri observes the same rule : the sixth one in order, al-Hakim,
was killed by his sister, the twelfth one, az-Zafir. was deposed and slain.
Then there follows the dpiasty of the Banu A\yub in Egj-pt ; the
sixth one of them, al-'Adil as-Saghir, was seized and deposed by the
nobles ; he was succeeded by only two members of the dynasty.
The rulers of the 'Ubaydi and the following Turkish (Mamlfik) dynasty
are recorded by ad-Damiri twice : once on the authority of Salahaddin
as-Safadi's Shark Lami'at al-'Ajam, in which the Mamluks are
enumerated only as far as al-Mansur Qala'fm al-Alifi.- and another
time in a detailed account of the 'Ubaydi dynasty based on authorities
different from Salahaddin as-Safadi's work, followed by a more com-
plete list of the Mamluks down to Sultan Murad.® In this list, too,
ad-Damiri points out that every sixth ruler was deposed or slain, thus
al-Muzaffar Qutuz who was slain, then al-Qahir Baydar who after
remaining in power for only half a day was slain, then al-Ashraf
Kujuk who was first deposed and then slain, then al-Malik as-Salih
Sahh who was deposed and imprisoned, then Faraj who was deposed
and slain ; the succeeding rulers of the Turkish djuiasty were nearly
all deposed or slain.
Throughout his History of the Caliphate ad-Damiri sets great store
by recording how the caliphs, especially the sixth ones, ended their
lives. In describing this he never fails to mention the cases in which
the deaths of the caliphs were foreboded by evil omens. Thus, speaking
of ‘Umar al-Fariiq's death, he mentions in the article ad-dlk (the cock) *
that 'Umar once dreamt that he had been pecked by a cock. When he
related his dream to Asma bint Umays she said that one of the
foreigners would slay him, as, in fact, happened a few days later when
> Ibid., p. 90, 11. 25-6. = Ibid., p. 94, 11. 23-32.
3 Ibid., p. 94, 1. 32, to p. 96, 1. 12. Ibid., p. 313, U. 23-31.
148
JOSEPH DE SOMOGYI
he was assassinated by Abu Lu’lu'a, a slave of al-Mughira ibn Shu‘bad
Also the death of al-Husayn was foreboded to the Prophet, as is
referred to in the article al-kalb (the dog) : The Prophet once dreamt
that a particoloured dog had thrust its muzzle into his blood and lapped
it, and he had explained it as meaning that a man would kill his
daughter's son al-Husa^m. Now ash-Shaniir ibn dhil-Jawshan, the
murderer of al-Husa^m, was the particoloured dog for he was leprous,
and the dream thus remained to be accomplished fifty years after the
Prophet. 2 Similarly we read in the biography of the founder of the
‘Abbasid dynasty, Abul-'Abbas as-Saffah ® on the authority of Ibn
Khallikan that Abul-'Abbas one day heard a slave say to another
slave : “ The term between me and you is two months and five days,'’
from which words he augured a bad omen, and he actually died of
small-pox two months and five days later.
II
The History of the Caliphate in the article al-vwazz is a concise
compendium on this subject written with the object of proving the
theory that every sixth ruler of Islam was either deposed or abdicated.
Ad-Damiri apparently did not intend to extend his history, as is
evident not only from his concluding statement : “ We have certainlv
expatiated upon that (i.e. upon the History of the Caliphate) but it is
not devoid of benefit or benefits,” < but also from the numerous
references in his History to other articles of the Haydt al-hayawdn,
in which he treats of a historical event or person in more detail. That
zoology is but the frame of the work can best be seen from the fact
that nearly every article of some length contains one or several
digressions of historical or biographical concern. Lea\-ing aside
ad-Damirl's very numerous biographies of scholars, theologians, and
poets, and confining ourselves to the historical matter in the strict
sense of the word, we find throughout the Hayat al-hayawdn frequent
digressions of this sort, which supplement the historical material in
the article al-iwazz. These digressions, to which references are in many
cases made in the history of the Caliphate, extend over all periods of
Islamic history, and their connection with the subject-matter of the
1 According to a saying of ‘A'isha quoterl in the article as-sabantM (a bold leopard,
vol. ii, p. 14, 11. 26-7), the genii wailed for ‘Umar three days before his death.
2 See vol. i, p. 56, 11. 16-25, and vol. li, p. 245, 11. 14-17.
® Ibid., vol. i., p. 69, II. 5-10.
* Ibid., p. 96, 1. 12.
A HISTORY OF THE CALIPHATE
149
articles in wliicli they are included is as loose as that of the History
of the Caliphate with the description of the goose.
As might be expected from an author so many-sided and so well
versed in all branches of Arabic literature, ad-Damiri drew on a number
of reference-works for both his History of the Caliphate and his
digressions in the other articles of the work, ilany of his sources are
well known by their printed editions, as is the case with his most
frequently quoted reference-work for history and biography. Ibn
Khalhkan's Kitdb wajaydt al-a‘ydn tea anbaabna az-zamdn. which,
by the English translation of MacGuckin de Slane. has been made
accessible to the general reader also. Ad-Damirl's constant references
to this biographical dictionary account for the prevalence of the
biographical over the purely historical matter in his work. Like so
many of his illustrious predecessors, he also lays more stress on
narrating the genealogy and the personal qualities of the rulers and
the interesting episodes of their lives than on recording their wars
and conquests.
The many-sided knowledge of ad-Damiri and the principal feature
of his HayCit al-hayawan as well, may best be seen from a STOoptic
table showing (1) the authorities quoted in the History of the Caliphate,
(2) the articles other than al-umzz referring to the history of the caliphs,
and indicating volume, page, and lines containing the narrative,
(3) the authorities quoted in these latter.
Authorities sub
Articles other than
Authoritk.’i of these
Caliph.
abiwazz. al-iwazz.
(1) The PRE-ISLAjnc Period
articles.
Destruction of
al-hhxiJd (the mole>
Ibn ‘Abbas, Wahb
the dam of
rat), i, 270, 26,
ibn Mnnabbih,
Ma‘rib.
to 272, 4.
Ibn al-Jauzi,
Qatada, as-
S u h a y 1 i , a 1 •
Mas‘udl, ash-
Sha‘bl. Abu
Sabra an-
Nakha'i.
Expedition of
al-Jil (the ele-
Su n an a^-Sahdh
Abraha, King
phant),, ii, 200,
of Abu 'AH ibn
of Ethiopia.
16, to 202, 31.
as-8akan, •‘^Tra
of Ibn Hisham,
as-Suhayll,
Sahlk of al-
Biikhari, Suvan
of Abii Da'ud
and an-Xasa'i,
Abu Walld al-
Azraql, a poem
by ad-Damiri
from the Kitdb
as-siyar^ part i.
1
150
JOSEPH DE SOMOGYI
Caliph.
Story of Jadhima
al-Abrash,
King of al-Hira,
and az-Zabba’,
daughter to
Mulayh ibn al-
Fara’.
Authorities sub
al-iwazz.
Articles other than
al-iwazz.
al - ' uqdb (the
eagle), ii, 1 13,
35, to 117, 15.
Authorities of these
articles.
Ibn Hisham, Ibn
al-Kalbr, Ibn
al-JauzI, the
commentator
of Ad-Duray-
diyya, af-Ta-
bari, Ibn as-
Sikklt.
Muhammad.
(2) MUHAinlAD AND HiS SUCCESSORS
(Battle of XJhud.)
Abu Bakr.
‘Umar.
‘Uthman.
‘All.
al-Hasan. ^
Zayn al-‘AbidIn,
grandson of
‘All.
Ibn Qutayba.
al-Fada’ili.
Ibn Qutayba, the
Imam Ahmad,
at-Tirmidhi,
Ibn Mahdl, al-
Mada’ini, al-
Mahdawi, al-
Waqidl, Zubayr
ibn Bakkar,
Ibn Ishaq.
Manaqib of the
Imam Ahmad,
Durrat al-yhaw-
u’ds,'- at-Tabari.
ash-Sha‘bI.
al-Burdq, i, 108, 7,
to 109, 11.
al-‘dtiq (the young
of birds), ii, 95,
25, to 96, 2.
al~arada (the
wood-fretter), i,
19, 4-16.
ash-shi'rd (species
of flies), ii, 45,
3, to 46, 17.
al-birdhaun (the
jade), i, 109, 31,
to 110, 4.
al-jamal (the
camel), i, 1 80,
24, to 181, 27.
al-baqhl (the mule),
i. i27, 4, to 128,
4.
Mu'jam of Ibn
Qani.
The two Sahihs,
Ibn Sa‘d, Sunan
of Ibn Maja.
Sira of Ibn
Hisham.
Kamil (of Ibn al-
Athir ?).
Ibn al-Athir, ad-
Dabbi, Ibn
Khallikan, al-
Hakim, Ibn abl
fihayba, Ibn
Dihya.
Kamil of al-Mu-
barrad, Kabi'al-
abrdr of az-
Zamakhshari,
Ibn ‘Asakir, az-
Zuhrl, al-
Asma'i, Ibn al-
Fallas, a 1 -
Mada'ini.
(
Mu'awiya.
(3) The Umayyads
al-himdr al-ahli Ibn Khallikan and
(the domestic others,
ass), i, 224, 14,
to 225, 1.
* By either al-Gliazali or al-Hariri : both the «orks were consulted bj- ad-DamIrl. »•
See my “ Index ”, Ko.s. 295 (al-tlhazali) and 432 (al-Harirl). •
A HISTORY OP THE CALIPHATE 151
Caliph.
Yazid ibn Mu-
‘awiya.
Marwan ibn al-
Hakam.
‘Abdalmalik ibn
Marwan.
(Hajjaj.)
al-Walid ,ibn
‘Abdalmalik.
(Musi'i ibn Nusayr )
Sulayman ibn
‘Abdalmalik.
‘Umar ibn ‘Ab-
dalaziz.
Yazid ibn ‘Ab-
dalmalik.
Hisham ibn ‘Ab-
dalmalik.
al-Walid ibn
Yazid.
Yazid ibn al-
Walid.
Marwan ibn Mu-
hammad.
Authorities sub
al-iwazz.
Al-akhbar at-thml
of Abu Hanifa.
Al-mustadrak ‘aid
's-sahihayn of
al-Hakim.
Kitab al-mahdsin
ival-niasdwl of
Ibrahim ibn
Muhammad al-
Bayhaqi, Al-
akhbdr at-tiwdl
of Abu Hanifa.
Ibrahim ibn abi
‘Abla, Ibn ‘Asa-
kir, ‘Alqama ibn
Safwan.
Kamil of al-Mu-
barrad, Muham-
mad ibn Sirin,
Al-mufndiial.
the Imam Ahmad,
Tabaqdt of Ibn
Sa‘d, Muham-
mad al-Mar-
■nazi, Raja ibn
Hay at, Ibn
‘ Asakir, Sirat al-
‘ CmarayU; Al-
hilya, ash-Sha-
fi’i.
Ibn ‘Asakir.
Mu-s'ab az-Zu-
bayri.
Ibn "Asakir, Kitdh
ailab wi-din ua
d-dunyd of al-
Mawardi.
ash-Shafi‘i.
Articles other than
al-iwazz.
ash-shat (the sheep
or goat), ii, 37,
16, to 38, 7.
al-fahd (the lynx),
ii, 19.5, 35, to
196, 21.
at -toys (the male
of the goat
kind), i, lUS, 3,
to 156, 25.
al-labu'a (the lion-
ess). ii, 273, 34
to 275, 24.
at-td'ir (the bird),
ii, 80, 12-27.
ad-dajjnj (the hen),
i, 299, 35, to
300, 11.
al-jarad (the
locust), i, 170,
26, to 171, 1.
al-hirr (the cat),
ii, 335, 3-16.
Authorities of these
articles.
Ibn KhaUikan.
al-Kiya al-Har-
rasi, al-Ghazali.
Muruj adh-dhahab
of aI-Mas‘udi,
Sharh as-Sira of
Qutbaddin,
Ta’rikh of Ibn
Khallikiin, and
others.
‘Aun ibn Shaddad.
Ibn Khallikan.
Ta'iikh of Ibn
‘Asakir.
Sunan of Abu
Da'ud, Ibn
‘Asakir.
Kitdh al-adhkiyd
of Ibn al-Jauzi.
Ibn Khallikan and
others.
(4) The ‘.^bbasihs
Abul-‘ Abbas as- Ibn al-Jauzi. al-birdhaun (the
Saffah. jade), i, 110. 27,
to 111. 21.
(Abu Muslim al-asad (the lion),
al-Khurasani.) i, 1, d, to 8, 14.
JOSEPH DE SOMOGYI —
152
Caliph.
Ja'far al-Man^ur.
Muhammad al-
Mahdl.
Harun ar-RashTd.
Muiammad al-
Amin.
‘Abdallah al-Ma’-
mun.
al-Mu‘ta?im.
Harun al-Wathiq
billJh.
Ja'far al-Muh-
tadi billah ibn
Harun.
Authorities sub
al-iwazz.
Ibn al-AthIr, Mu-
hammad ibn
Zafar.
Al-akhbdr at-iitml
of Abu Hanifa,
al - Asma‘i,
‘Uyun at-ta-
loarikh (of al-
Kutubi).
Al-akhbdr at-iiu’dl
of Abu Hanifa.
M a jma' al-akhbdr
of al-‘lraql, ash-
Shafi‘I, al-Ja’iza,
Ibrahim al-
Harbl, Ahmad
ibn Sinan, ‘Abd-
allah ibn al-
Ward, Tahdhib
al-asmd wal-
lu'jhdt of an-
Nawawl, Mu-
hammad ibn
Khuzayma.
Duwal al-isldm of
adh-Dhahabi,
Hilya al-anbiyd
xm tahaqdt al-
asfiyd of Abu
Nu'aym.
Muhammad ibn
al-Husayn ibn
‘Abdallah al-
Baghdadl.
Articles other than
al-iwazz.
al-bdzl (the hawk),
i, 101, 18, to
102 , 6 .
al-baghl (the
mule), i, 132, 5,
to 133, 31.
al-bdzi (the hawk),
i, 99, 26, to 100,
10 .
al-ba‘ ir (the
camel), i, 123,
22, to 125, 17.
al-baghl (the
mule), i, 129,
11, to 130, 14.
al-‘ugdb (the
eagle), ii, 11],
23, to 113, 12.
a I -fa ra « (t h e
mare), ii, 187,
22, to 188, 33.
al-fahd (the lynx),
ii, 196, 28, to
197, 4.
al-birdhaun (the
jade), i, 110, 4-
27. ■
Authorities of these
articles.
Nahw of Siba-
wayh.
Kitdb al-adhkiyd
of Ibn al-JauzI.
Ibn Khallikan.
Sirdj al-muluk of
at Turtushl, Ibn
Balban, Sharh
al-asmd al-
husnd of al-
Maqdisl, and
others.
Ta'rikh Baghdad,
Ibn KhaUikan.
Ibn Khallikan, Ta-
'rikh Sahib Ha-
mdt, and others.
Ibn Balban, al-
Ghazall, and
others.
Ibn Khallikan.
Ibn Khallikan.
A HISTORY OF THE CALIPHATE 153
Caliph.
al-Muktafi billah
ibn al-Mu‘ta-
did.
Abul-Fadl Ja‘far
al-Muqtadir
billah.
(al-Hallaj.)
Muhammad al-
Qfihir hillah.
‘Abdalkarim at-
Ta’i billah.
‘Abdallah al-Qa-
’im bi amr
Allah.
Buyids : —
(a) ‘Imad ad-
daula ibn Bu-
wayh.
(6) Rukn
ad-
daula.
(c) ‘Adud
ad-
daula.
The w h
o 1 e
dynasty of the
Fatimids.
al-Hakim al-
‘ijbaydi.
Nuraddin.
Authorities sub
al-iuazz.
the books of adh-
Dhahabi.^
An-nashwdn of
at-Tanukhi.
Ta’rikh of Ibn al-
Bitriq.
Ba's mdl an-nadim
of Ibn al-Bitrlq,
Ibn al-AnbSri.
Ta’ril-h of Ibn al-
Bitriq.
Sharh Ldmi''at al-
‘ajam of Sala-
haddin a§-Sa-
fadi.
Articles other than
al-iwazz.
al-himdr al-ahll
(the domestic
ass), i, 223, 14,
to 224, 14.
al-hayya (the ser-
pent), i, 2.o6,
6 - 20 .
ad-ddbba (the
beast), i, 287,
26, to 288, 3.
al-fahd (the Ivnx),
ii, 197, 1-9.
al-himdr al-ahli
(the domestic
ass), i, 227, 29,
to 228, 18.
al-batt (the duck),
i, 114, 8, to 116,
19.
Authorities of these
articles.
Ibn Khallikan,
jBafdlih al
kunuz of al-
Maqdisi. Mish-
qdt al-anicar wa
misfut al-a'irdr
of al-Ghazall,
and others.
Ibn Khallikan
from the Ta’-
ril h of Abu Mu-
hammad Harun
ibn al-'Abbas al-
Ma'muni.
Ibn Khallikan.
Ibn Khallikan.
Ibn Khallikan,
Ibn al-Athir.
(5) The Latek Dynasties
' How much adh-Dhahabi's works were appreciated by our author is evident
from his record on the caliphate of Muhammad al-Muqtafi h amr Allah : that
according to “ our friend the Hafiz Salahaddin Khalil ibn Muhammad al-Aqfahsi ”
this caliph was followed by a certain al-Mustazhir but, not knowing this, he does not
mention him, the more so as adh-Dhahabi mentions the caliphs in the order as he
gives them (see i, 89, 4-6). A similar discordant view of adh-Dhahabi is also mentioned
concerning al-Hakim bi amr Allah Ahmad ibn al-Mustakfi billah who, according to
al-Husayni in his Dhatjl ‘aid l-‘lbar, followed his father al-Mustakfi billah, whereas
adh-Dhahabi asserts at the end of his l>haijl that he was followed by his brother
Ibrahim (see vol. i, p. 91, 11. .30—4).
154
JOSEPH HE SOIIOGYI—
Caliph.
Abul-'Abbas Ah-
mad ibn Tulun.
Malikshah as-
Saljuql.
Mahmud ibn Se-
buktekln.
Authorities sub
al-iwazz.
Articles other than
al-iwazz.
ad-dajjdj (the
fowl), i, 299, 19-
3.5.
al-haqar al-ahll
(the domestic
cow), i, 137, 1-
20 .
al-qumrl (the ring-
dove), ii, 225,
7-27.
Authorities of these
articles.
Kitdb al-adhkiyd
of Ibn al-Jauzi,
Ibn KhaUikan.
Ibn KhaUikan.
Ibn KhaUikan,
Ibn al-Athir.
The names mentioned in the above table do not exhaust all the
authorities consulted by ad-DamIrl, because in many cases he gives
such general statements as “ historians relate ” or “ so men say ” or
“ others say But even from the sources mentioned by him we may
infer that he made use of practically all the sources available for him.
Thus the Hayat al-hayawan is also a good guide to the Arabic authorities
on general history.
The History of the Cahphate, itself a digression in the article
al-iwazz, also contains two digressions. The one is inserted into the
subchapter on the caliphate of ‘All and, under the title of Fa ida
Intlfa (“ an agreeable item of information ”)} contains a list of the
names of the prophets, the nine persons who used to write for the
Prophet, the nine collectors of the Qur'an in His time, the four
executioners in His time. His five bodyguards and fourteen com-
panions, the seven tdbi‘s of al-Madina whose legal decisions were
final, the four persons who. according to tradition, talked in their
cradles, the four persons who, according to tradition, talked after
their deaths, the four persons whom their mothers carried in their
wombs beyond the usual period of pregnancy, the six Ximrods, the
three Pharaohs mentioned in tradition, the founders of the five
religious schools (ma lJihahs) ~ and the dates on which thev died on
the authority of an-XawawTs Kildh ‘ulum al-ha llth, the six relaters
of tru.stworthy traditions ^ and the dates on which they died. This
list testifies to ad-Damlrl's fondness of curio.sities, and his intention
of collecting in his book information of various kinds which might
be useful to a Muslim reader. The .second digression is inserted in
the sub-chapter on the caliphate of al-Musta'In bilhlh,^ and gives
‘ Vol. i, p. 51, 1. 18, to n. 52, I. 24.
- In addition to the four orthodox schools that of .Sufyan ath-Thaurl (lived in the
eighth century a.d.) is also mentioned by ail-l).miirl.
^ Instead of Ibn M.’ija. ad-DamIri quotesad-Ddraciiitmaa the sixth one.
' \’ol. i, p. 93, 1. 3, to p. 94, 1. 14.
A HISTOEY OP THE CALIPHATE 155
maxims on the right conduct to be observed by persons who have
intercourse with caliphs, kings, and sultans, and expounds the
principles of morality and the good qualities of people. The treatise
is full of proverbs and parallelisms like all works of instruction in
Arabic, and Oriental literature in general, and refers to such authorities
as Iflatun (Plato), ash-Sha'bl, al-Fadl ibn ar-Rabi‘, Khalid ibn
Safwan, ash-Shi'b of al-BayhaqT, the Imam Ahmad, and others. ^
^ For a similar treatise on polite manners see also the article as-saqr (the hawk),
vol. ii, p. 58, 1. 2, to p. 59, 1. 14.
4
Nadir Shah’s Campaigns in ‘Oman, 1737-1744
By Laurence Lockhart
'IITIRZA 3LAHDI, the biographer of Nadir Shah, gives but little
information on the subject of the Iranian campaigns in ‘Oman
during the period 1737 to 1744, probably because ‘Oman was, relatively
speaking, merely a secondary theatre of war and because Nadir did
not go there in person. Fasa'I, in his Farsnamu-yi-yasirl, also devotes
but little space to the subject, notwithstanding the fact that
Muhammad Taqi Khan ShlrazI, the Beglarbegi of Fars, played a
prominent part in the campaigns.
Arabic and European sources, on the other hand, provide a con-
siderable amount of data. Of the contemporary authorities, the most
important are the records of the representatives of the East India
Company in the Persian Gulf ; these records have not, so far as I am
aware, yet been utilized at all by anyone who has written an account
of the campaigns. J. Otter, who was stationed at Basra for part of
the period, makes some reference to the expeditions in his Voyage en
Turquie et en Perse, ^ but his information is neither very full nor very
accurate ; of decidedly more value is Carsten Niebuhr's account of the
geography and history of ‘Oman in his Beschreibung ron Arahien,'^
but Niebuhr was not, strictly speaking, a contemporary.
In 1856 appeared Charles Guillain’s ® Documents sur I'Histoire,
la Geographic et le Co^nmerce de I'Afrique Orientale, in which, in the
section relating to the ‘Omani connection with East Africa, some
valuable information is given ; this information is derived mainly
from the Arabic manuscript of a certain Shaikh Abu Sulaiman
Muhammad ibn Amir ibn Rashid. In the same year. No. xxiv of the
Brymbay Government Selections was published, which contained,
mter alia, Captain Robert Taylor’s “ Extracts from Brief Notes
containing Historical and other Information respecting the Province
of ‘Oman ’ ’ and Francis AVarden's “ Brief Notes relating to the Province
of ‘Oman ” ; neither of these writers gives any data of value respecting
^ Published in Paris in 1748.
2 The references to this work which are given below are to the Copenhagen edition
of 1772.
® Guillain, a French naval officer, visited Zanzibar in 1846.
158
LACREXCE LOCKHART
the ‘Oman campaigns, but some of Captain Taylor's geographical
notes on Khor Fakkan and Julfar (Rasu'I-Khaima) are of interest.
The principal work in English on this subject is The History of the
Imams and Seyyids of ‘Oman ", which the Rev. G. P. Badger trans-
lated from the Arabic of Salil ibn Razlq ; this work was published by
the Haklmd Society in 1871. Salll ibn Razlq gives in great detail the
history of the Iranian operations in ‘Oman, but his statements must
be accepted with reserve, because he was much addicted to
exaggeration ; moreover, but little reliance can be placed upon his
chronology.
Of more recent writers, mention may be made of ; — •
(1) Lieut.-Colonel E. C. Ross, whose “ Outlines of the History
of ‘Oman from 1728-1883 ” was included in his Report on the
Administration of the Persian Gidf Residency and Muscat Political
Agency, 1882-3 ; this article is based on Salil ibn Raziq's history.
(2) Colonel S. B. Miles, whose book, The Countries and Tribes
of the Persian Gulf, was published in 1919 ; as regards the Iranian
campaign, Colonel Miles relates nothing which is not in Badger’s
translation of Salil ibn Razlq.
(3) Clement Huart, in his Ilistoire des Arabes, gives a brief
outline of the ‘Oman campaign.s. but his account is merely a
paraphrase of Salil ibn Razlq.
(4) A. Grohmann makes a brief reference to these campaigns
in his article “ 31askat ” in volume iii of the Encyclopsedia of
Islam. He appears to have relied mainly on Salil ibn Razlq,
and his chronology con.sequently needs revision.
(5) R. Said-Ruete. Though this vTitcr'.s .special study is the
Al-Bu Sa‘Id Dynasty, 1 he has also written of the immediately
preceding period in ‘Oman, and gives in detail the story of how
Ahmad ibn Sa‘Id rose to power. He has made a careful study of the
authorities available, with the exception of the records of the East
India Company (of the existence of which he does not seem to have
been aware).
It is the purpose of the present uviter to attempt to correlate the
information contained in the records of the East India Company’s
representatives with the data given by Salil ibn Razlq, Otter, Niebuhr,
Guillain, and certain other authorities.
^ See his Said Bin Sultan^ 1791-1856, London 1929, and The Al-bu-Said Dynasty
in Arabia and East Africa in the Journal of the Central Asian Society, vol. xri, part iv.
160
LAURENCE LOCKHART —
In the reign of the gentle, but incapable, Shah Sultan Husain
(1694^1722), Iranian prestige and influence in the Persian Gulf suffered
greatly, while there was a corresponding growth in the power of the
Arab tribes inhabiting the coastal strips on both sides of the Gulf.
The Arabs of Muscat, under the Imam Sultan ibn Saif II (1711-
1718) had a powerful fleet,i with the aid of which they made, in 1717
or the following year,^ a successful descent upon the Bahrain Islands ;
these islands had, for nearly a century formed part of the Iranian
Empire. The ‘Omani Arabs do not appear to have remained long in
Ba^ain which, on their departure, passed into the hands of Shaikh
Jabbara, the chief of the powerful Huwala Arabs. Though nominally
an Iranian subject, Shaikh Jabbara was, to all intents and purposes,
independent of the weak Government at Isfahan.
When Nadir Shah assumed control of the Government of Iran. ®
the state of affairs which he found prevailing in the Persian GiJf
proved intolerable to him. He took active measures to reduce the
coastal Arabs to submission, but this task, as he soon found to his cost,
could never be completed until he had a fleet of his own. Nadir set
about the formation of his fleet by effecting the purchase (more often
by forced sale than not) of several vessels from the English and Dutch
East India Companies.
The first venture of the nascent fleet was an unsuccessful attempt
to capture Basra in May, 1735.* In the following year, however,
Latif Khan, the Iranian Daria-Begi, or Admiral, succeeded in wrestini^
Ba^ain from the Huwala Arabs, when the redoubtable Shaikh
Jabbara was absent on a pilgrimage to Mecca.®
Having obtained the upper hand over the Huwalas, Nadir,
apparently at the suggestion of his Admiral, Laflf Khan, determined
> C.i|itain Alexander Hamilton, in his work, A Xew Account of the East Indies,
Edinburgh, 1727, vol. i, p. 74, gives interesting particulars of the formidable fleet
which the Imam of ‘Oman possessed in 1715.
2 The exact date is uncertain.
3 strictly speaking, it is not correct to refer to him as “ Nadir ” until his accession
in 1730. He was originally known as Nadr or Nadhr Quli Beg ; in 1726 Tahmasp
Shah gave him the title of Tahmasp Quli Khan, which he held until he became Regent
in 1732, when he was known officially as Wakllu’d-Daula and Na’ibu’s-Saltana.
‘ detailed description of this attack on Ba-sra is given by Martin French, the
representative of the East India Company in that town, in an interesting letter which
he wrote to the London office of his Company on the 5th/16th June, 173.5 ; this letter
Ls contained in volume xv of the Factonj Records of Persia and the Persian Gulf in
the India Oflice.
^ See Mirza Mahdi’s TaWikh-i-Xadiri. Borabav, 1849, p. 172, and p 113 of
At-Tuhfatv'n-Xabhdniyya fi TarUkil-Jazirati'l.‘Arabiyya (Cairo, 1929-1930) bv
Muhammad ibnu’sh-Shaikh Khalifa. ’ ^
NADIE shah's campaigns IN ‘OMAN
161
to carry out a more ambitious project, namely tbe conquest of ‘Oman.
Nadir doubtless felt that, until Muscat and tbe other ports on the
Arabian coast were in his hands, he coidd not hope to establish and
maintain Iranian supremacy in the Gulf.
Fortunately for Nadir, events in ‘Oman took, in 1736, an
unexpectedly favourable turn. The licentious ways of the Imam
Saif ibn Sultan having aroused the fierce resentment of the more
rigidly orthodox of his subjects, a serious revolt broke out in 1736.
Saif tried in vain to stem the growing tide of opposition by importing
mercenaries from Makran.^ Faced with defeat and deposition, Saif,
in despair, turned to Nadir Shah for aid. Although already in the throes
of preparation for his onslaught on the Ghilza'i Afghans, Nadir was
quick to seize his opportunity. He sent prompt orders to Muhammad
Taqi Khan ShirazI, the Beglarbegi of Fars, to equip a fleet and landing
force, and to go to ‘Oman ostensibly to help the hard-pressed Imam
but in reality to take possession of his country. IVhen, in January,
1737, a report of Nadir’s intention to send an expeditionary force to
‘Oman reached the Agent of the East India Company at Gombroon,
he informed the London oflSce that he hoped that the Shah was not
intending to deceive the Imam and seize his country ; he feared that
the attempt would be unsuccessful and that it would merely result in
the Imam taking retaliatory measures.^
On the 3rd/14th March, 1737, the Iranian fleet arrived at Gombroon
from Bushire, under the command of Latif Khan, “ who hoists his
flag, being a white ground with a red Persian Sword in the middle.” ®
The fleet consisted of four ships, two “ grabs and some smaller
vessels.
Latif Khan’s flagship, the Fath-i-ShdJi, had formerly been known
as the Cowan ; the Iranians had purchased her from the East India
Company in the previous December. Another of his large vessels, the
Northumberland, had also been British ; she had been acquired by
^ 8ee p. 132 of the Rev. G. P. Badger's translation of Salll ibn Raziq's history.
2 8ee the letter from the Agent of the East India Company at Gombroon to London,
dated the 28th JamiarjySth February, 1737 ; this letter is in vol. xv of the Factory
RecordSf Persia and the Persian Ovif.
® Letter from Gombroon to London, dated 6th/17th May, 1737, in vol. xv of the
Factory Records.
^ An English corruption of the Arabic word ghurab^ Trhich means ordinarily a
“ crow ”, but which was also used to denote a type of vessel much in use by the Arabs.
Olaf Toren, in his book. A Voyage to Suratte, Chinas etc. (London, 1771), p. 205, describes
a “ grab ” as a two- or three-masted vessel, rigged in the European manner, with a
low and sharp prow, like that of a galley.
VOL. VIII. PART 1.
11
162
LAX.TIEXCE LOCKHART —
forced sale at Bushire nearly a year befored The captain (named
Cook) and the mate of a British vessel, who had entered the Iranian
service, each commanded an Iranian ship. The crews consisted entirely
of Huwalas and other Arabs from the Gulf Coast.
Having embarked 5,000 men and 1,500 horses, the fleet sailed
from Gombroon on the lst/12th April ^ and reached Khor Fakkan,
74 miles south of Has Musandam, fom days later.® Latif Khan landed
some of his Iranian troops at Khor Fakkan and then sailed back
northwards, rounded Eas Musandam, and anchored in the harbour
of Julfar ; Latif Khan here met the Imam Saif ibn Sultan who, it is
said, made the Admiral a present of 2,000 tomans (some £4,400).*
After recemng the submission of Shaikh Eama, the chief of the local
(Huwala) Arabs, and leaving a garrison in JuKar, Latif Khan and his
forces, accompanied by Saif and his adherents, marched inland to
meet Bal'arab ibn Himyar Al-Ya‘riba who, although a relative of the
Imam’s, was the leader of the opposing faction. Bal'arab, for his part,
was advancing northwards from Nizwa. A battle took place at Falju's-
Samini, near Biraimi, which resulted in a decisive victory for the allies,®
who thereupon occupied successively Al-Jauf and ‘Ibra.® The advance
was not continued, however, because dissension had broken out
between Latif Khan and the Imam ; it appears that the former was
already acting as though the Iranians had conquered the whole country,
and was not disposed to treat Saif with any deference.’ In consequence,
the alliance between the Imam and the Iranians was, for a time,
dissolved ; in July Latif Khan returned to Gombroon, where the
town was illuminated and feasts were held in his honour.®
^ This vessel did not, apparently, belong to the East India Company.
2 See the Gombroon letter to London of the 6th/] 7th May, to which reference
has already been made. See ako J. A. Saldanha’s Selections Jrom State Papers^ Bombay,
regarding the East India Company's mnn^rlion v'ith the Persian Gulf v:ith a summary
of events, 1600-1800, Calcutta, 1908, p. 53.
® Guillain, in vol. i, j). ij'ZU, of Ids Documents, gives (on the authority of Shaikh
Abu Sulairaan) the date of the arrival of this force at Khor Eakkan as the 14th Dhu’l-
Hijja, 1149 (16th April, 1737); Sabll ibn Razlq gives the date as the 17th Dhu'I-
Hijja (see Badger, p. 13S) ; as Shaikh Abu Sulaiman’s chronology is far more accurate
than that of Salil, I have adopted the former’s date, which, murcfiver, tits in better
with the date of departure of the fleet as given by the Gombroon Diary.
^ See the entry, dated 15th/26th May, in the diary of the Agent-in-Council of
the East India Company ai Gombroon, in vol. iv of the Factory Records.
® See Guillain (again on the authority of Shaikh Abu Sulaiman), vol. i, p. .529,
and Badger, p. 141.
® Salil sa\s (see Badger, pp. 141 and 142) that the Iranians behaved in a most
barbarous manner to the inhabitants of ‘Ibra.
’ Guillain, vol. i, p. 529.
® Gombroon Diary, 21.st July/lst August, 1737.
NADIR shah’s campaigns IN ‘OMAN
163
Taqi Khan, the Beglarbegi of Fars, who had previously arrived at
Gombroon, took over the supreme command from Latif Khan, and
by his parsimony caused much discontent amongst the seamen. Being
unpaid and half-starved, the seamen complained bitterly to the
Admiral, who promised to redress their grievances as soon as the
Beglarbegi departed. Operations were not resumed during the summer
and autumn, and in Kovember the Iranian fleet underwent repairs
and refitting at Gombroon. It appears that at about this time Nadir
Shah 1 sent the Beglarbegi peremptory orders to prosecute the ‘Oman
campaign with vigour, and upbraided him for not leading the
expeditionary force in person in the previous April.
Spurred to action by his sovereign's censure, Taqi Khan proceeded
to commandeer all the English and Dutch trankeys at Gombroon,
and importuned both the Companies for the loan of a large vessel
to supplement his fleet. By means of a heavy bribe, the Agent of the
English Company managed to avoid lending him a ship which was
already loaded with goods and on the point of sailing for India.^
The Beglarbegi then concentrated on the Dutch, with whom he had a
violent quarrel ; it is not, apparently, recorded whether in the end he
obtained one of their ships. ^
In January, 1738, Taqi Khan, in company with Latif Khan and
the Khan of Lar, sailed for Julfar with some 6,000 men. In the
following month word was received at Gombroon that “ the
Beglerbeggy, who imagined that He should cajole the Imaum into his
Power and by that means make himself Master of Muscatt, has
quarrelled with Lettif Cawn for advising the King of this Enterprise,
in which he begins to fear he shall meet with no great success, being
himself no Soldier . . ." The Beglarbegi and Admiral, nevertheless,
smoothed over their differences for a time, and joined forces with the
Imam, whose lack of success against his rebellious subjects had forced
him once more to seek the aid of the Iranians. In March, 1738, the
combined forces again defeated Bal'arab ibn Himyar and occupied
the towns of Bahia and Nizwa ® ; they then advanced to Muscat,
* Letter from Gombroon to Lomlon, dated 1.5th/26th December, 1737 (in vol. xv
of Factory Eecords).
- Gombroon Diar}', 14tti/25th December, 1737.
’ I have had no opportunity of consulting the records of the Dutch East India
Company.
* Gombroon Diarv", 2nd/13th February, 1738.
' According to Salll ibn Raziq (Badger, p. 142), the Iranians captured Bahia on
the 14th March and Xizwa eight days later; the news reached Gombroon on tlie
8th/19th April, uhen celelirations were held in honour of the occasion. See also
164
LAXTRENCE LOCKHAET
whicli they occupied without difficulty, but the Iranians were unable
to secure possession of the eastern and western ports (known
respectively as Jalali and Maranl),^ although they are said to have
besieged them for five weeks.^ Soon after the occupation of the town
of Muscat, Taqi Khan quarrelled with the Imam Saif, with the result
that the latter withdrew with his fleet to Barka, whence he proceeded
into the interior and entered into negotiations with his rival, Bakarab
ibn Himyar. Bal'arab not only agreed to renounce all claim to the
Imamate, but promised to assist Saif against the Iranians.®
As TaqI Khan was unable to capture the Muscat forts, he raised
the siege of them on the 25th May, 1738, and went with his army to
Barka where, as at Muscat, the forts successfully resisted him. Here
dissensions once more broke out between Taqi Khan and Latif Khan,
in consequence of which the former poisoned the luckless Admiral.* It
seems that Taqi Khan was then reduced to great straits, and that he
was compelled to retreat to JuLfar.® He was able, however, to detach
a portion of his force to besiege the town of Sohar ; meanwhile, disaster
had overtaken the Iranian garrison at Bahia, which was besieged and
forced to surrender after the majority had been killed.® Later the
Iranians besieging Sohar were forced to give up their investment of that
place and to retire to Julfar.
MTiilst the Iranians in ‘Oman were sustaining these reverses,
dissatisfaction was rapidly increasing amongst the personnel of their
fleet, owing to lack of pay and insufficient rations ; in consequence,
many of the men deserted, and those who remained were in no mood
loyally to fight for the Iranians. The natural result was that the
Huwala Arabs and the ‘Omanis regained the command of the Gulf,
and piracy became rife. Suppbes for the Iranian garrison at Julfar
Guillain, vol. i, p. 529. When Wellsted was at Xizwa in November, 18.35, he went over
the fort there which, he said (see his Trnveh in Arabia, London, 1838, p. 121), “ in
the estimation of all the surrounding country, is impregnable ”. He found in the fort
“ a few old guns ... one bearing the name of Imam Saaf, and another that of Kouli
Khan, the Persian General who took Maskat ” (*ie). The Iranian gun may have borne
Nadir’s former title of Tahmasp Quli Khan.
1 These forts which were built by the Portuguese in the sixteenth centurj-, were
also known as the Qafatu’sh-Sharqivya and the Qaratu’l-Gharblyya respectively ;
A. W. Stiffe gives a description of them in an article in the Geographical Journal,
vol. X, 1897, pp. 609-612. See also Curzon’s Pernia, vol. ii, p. 440.
“ Badger, p. 143.
3 Ibid.
* Gombroon Diarj', 26th June/7th July, 1738.
* Gombroon Diary, 13th/24th July, 1739.
* Badger, p. 144.
NADIE shah's campaigns IN ‘oMAN
165
had, for a time, to be conveyed across the Gulf by English vessels.
In August, 1738, the East India Company's galley Rose was captured
by two Arab “ grabs ”, and the officers and men were threatened with
death if they carried any more provisions for the Iranians.^ Almost
simultaneously the Arabs descended in force upon Basidu, on Qishm
Island, plundered the town, and slaughtered all the Iranians there.
Bad news was also received in Gombroon from Bahrain, where, it was
reported, the Arabs were besieging the Iranians in the castle.^ Later
in the year the Arabs, as usual, fell out amongst themselves, and were
in consequence much weakened, and in January, 1739, the Iranian
fleet defeated the Arabs, whose Admiral was killed by an explosion.^
In March, 1739, TaqI Khan returned to Gombroon armed with
full powers to carry on the war against the Arabs and with “ positive
orders to level Muscatt to the ground No operations could, however,
be carried out in ‘Oman because, in the first place, TaqI Khan was
called away from Gombroon before he could sail, in order to suppress
a revolt in the Kuhgilu country,* and, later in the year, when that
revolt had been quelled, he received orders from Nadir to proceed
with his land and sea forces to Sind.* The disastrous ending of this
expedition resulted in Nadir summoning TaqI Khan to Nadirabad
to account for his conduct.''
For the next two years nothing could be accomplished on the
Arabian shore, beyond retaining a precarious hold on Julfar. A serious
mutiny in the Iranian navy in August, 1740, ended in the vessels being
taken away by their crews, some going to the Arabian shore and others
to the island of Qais, which was held by the Huwala Arabs.® It was
not until early in 1742 that the Iranians obtained the mastery over
the mutineers and rebel Arabs, some more large vessels having hy then
been obtained from Surat, through the East India Company. The
1 Gombroon Diarj-, 26th July/6th August.
- Ibid.
“ Letter from Gombroon to London, dated 31st March/llth April, 1739 (in vol. xv
of the Factor]! Records).
J Ibid.
“ Letter from Gombroon to London, dateel loth/26th May, it was stated : “ We
shall be glad if this or some other Aeeident prevents his (the Beglarbegi’.s) return to
carry on the Expedition against Museatt which must quite ruin this miserable place.”
‘ Td'rJkh-iSddirt, p. 211 ; see also Otter, vol. ii, p. 87.
■ According to the Gombroon Diary, .)tb/16th April, 1739, “ the greatest want of
Conduct imaginable appeared in this expedition.” >See also the Tarikh-i-Xadirl, p. 214,
in which, however, but little uiformation is given of this expedition.
* There are numerous entries in the Gombroon Diarj- respecting this mutinj- and
its consequences.
166
lAUBENCE LOCKHAKT —
mutineers, for their part, had suffered a serious loss when two of their
vessels were driven ashore in a gale.
Events in ‘Oman now once more favoured Iranian intervention.
Saif ibn Sulfan had again estranged many of his subjects by his mode
of living and his disregard for the precepts of the Qu'ran, and in
February, 1742, his opponents broke into open revolt. On the 16th
of that month they were strong enough to secure his deposition and to
bring about the election of his cousin Sultan ibn Murshld.^ Saif, as
before, turned to the Iranians for aid, and TaqI Khan sent a favourable
reply.
The Huwala Arabs, who had not submitted to the Iranians, ranged
themselves on the side of Sultan ibn Mmshid and his supporters,
but the Jidfar garrison marched out and in April decisively defeated
the Arabs at Khasab, killing Shaikh Rama and taking over 500
prisoners.’^ The Julfar garrison was strongly reinforced in June,
but it was not until the following November that Kalb ‘All Khan,®
Nadir's brother-in-law, who had been appointed Sardar of the
Garmsirat, or Commander-in-Chief of the Hot Countries, crossed over
to the Arabian shore. He was followed three weeks later by TaqI
Khan himself. At this time the Iranian fleet was strengthened by
the arrival of four new ships from Surat ; later it received a further
increase in strength when Saif handed over two vessels to his allies ;
further, another vessel was obtained from the French, by a forced sale.®
The Beglarbegi, on meeting Saif ibn Sultan at Julfar, concluded
a treaty with the ex-Imam whereby he undertook to restore him to the
Imamate on condition that he (Saif) would then recognize the
1 Shaikh Abu Sulaimfin (see Guillain. vol. i, p. 5.3.5) givr.s thi.s date as the 10th
Khu'l-Hijja, llo4, the equivalent of the I6lh Fehruarv, 1742. Salil ihn Razlq is
obviously wrong in saying (see Badger, p. 145) th.at this event took plaee in a.h. 1151
(A.D. 1738-0). It is clear from the Gombroon Diary that the revolution took place in
1742, because the .Agent receiveil a letter at the beginning of April in that year from
the Beglarbegi asking for ships to take troops across to Arabia to assist Saif ibn
Sultan, ■■ who has been driven out bj' rebels."
= Gombroon Diary, lhth/30th April, Otter, vol. ii. p. 169. TaqI Kh.in, on hearing
of this victory, sent an order to Gombroon for 1,000 tomans to be rai.sed from the
local Baninm, which sum was to be given to the .Tulfar troops a.s a bonus. The Bnnian^
were so dissatisfied at this high-handed measure that they threatened to leave
Gombroon (Gombroon Diary, 23rd July/3rd .Aiigu.st).
» Kalb ‘All Khrin’s father was Bfibri ‘All Beg, of Ablvard, one of whose daughters
was Nadir’s first wife and wa.s the mother of Rida Quit Mlrzil ; after the de.ath of thi.s
wife, Nadir married another daughter of Baba ‘All Beg’s, Gauhar Shad bv name, who
was the mother of Na^ru’Uah Mirza and Imam Quit Jllrza.
* Gombroon Diary, 20th Novemlrer/lst December.
= For particulars of the sale of this French vessel, see Otter, vol. ii, pp. 166-8
NADIE shah’s campaigns IN ‘OMAN
167
suzerainty of Iran.^ The allies then advanced against Sultan ibn
Murshid and his adherents.
While Kalb ‘All Khan, with a portion of the Iranian troops, laid
siege to Sohar, TaqI Khan and Saif ibn Sultan proceeded bv sea to
Muscat, which was held by partisans of the ex-Iniam. On landing,
Taqi Khan found, to his annoyance, that although his men were
allowed to wander freely about the town, they were denied access to
the Jalali and Marani forts. Taqi Khan, however, was a resourceful
man, and he had, it appears, expected some such behaviour on the
part of Saif ibn Sultan. Knowing the ex-Imam’s addiction to drink,
Taqi Khan had, it is said,^ brought a cask of Shiraz wine in his flagship.
When Saif invited Taqi Khan and some of his officers to a banquet
in the Marani fort, the Beglarbegi presented his host with the cask of
wine, with the result that Saif, as well as his principal officers, became
completely drunk. While Saif and his officers were lying insensible,
Taqi Khan obtained possession of the fort without difficulty or blood-
shed. There still remained the other fort before the Beglarbegi could
claim that Muscat was completely in his power. Taqi Khan thereupon
stole Saif's seal and affixed it to an order which he had written, in the
ex-Imam’s name, to the Commander of the Jalali fort, charging him
to admit the Iranian troops. The Arab commander, believing, as was
natural, that Saif liimself had written the order and affixed his seal
to it, opened the gates to Taqi Khan. When Saif ibn Sulfan recovered
his senses he found, to his dismay, that his allies were masters of
Muscat.®
It being useless to attempt to oust the Iranians from the forts,
Saif decided to continue his alliance with them until the overthrow
of the Imam Sultan ibn Murshid could be accomplished.
All this time the siege of Sohar had been in progress ; the defence
was carried on with great vigour and ability by the Governor, Ahmad
ibn Sa'id by name, who, as will be seen below, later founded the
Al-Bu Sa'id dynasty.
Taqi Khan and Saif ibn Sultan advanced against Sultan ibn
^ Guillain, vol. i, p. 536 ; Otter, vol. ii, p. 163.
2 Niebuhr's Besrhreibung von Arabien, p. 300.
^ Niebuhr's Beschreibung von Arabien^ pp. 300 and 301. Guillain {vol. i, p. 537)
gives a slightly different version. The Agent of the East India Company at Gombroon
received word from the Iranian Government on the 18th February/lst March, 1743,
that the Beglarbegi had captured Muscat, Doubts whether Taqi Khan had taken
Muscat by orthodox methods were held, for the Agent added that it was supposed
that the place had been taken “ by dealing underhandedly with the Imaum’s slaves
to deliver him the Forts ”.
168
LAtTRENCE LOCKHART
Murshld, who was then marching to the relief of Sohar. It appears
that Sultan ibn Murshld succeeded in making his way through the
hues of the besiegers and in joining Ahmad ibn Sahd. The Imam was,
however, killed almost immediately after, when leading a sortie d
Saif ibn Sultan was, it is said, much affected by the death of Sultan
ibn Murshid ; overcome with remorse for his conduct, which had
brought so much trouble upon his country, Saif abandoned his Iranian
allies and retired to Eastaq, where he died not long afterwards.^
Notwithstanding the death of Sultan ibn Murshld, Ahmad ibn
Sa‘Id continued to resist bravely. In March the hopes of the besieged
were raised by a serious rever.se which the Iranians sustained. Two
thousand of their men, when returning laden with booty from a
marauding expedition, camped in a valley for the night. No proper
guards were set, and an Arab force was therefore able to take them by
surprise ; only a few wounded men escaped to tell Kalb ‘All Khan
of the disaster.^
Some five weeks later reports reached Gombroon of a further and
even more serious reverse, but these were later contradicted.*
By July Ahmad found his supplies were nearly exhausted, and,
as he had no hope of relief, he deemed it expedient to capitulate on
honourable terms to his powerful opponents. §ohaT thus passed into
the hands of the Iranians, but it had held out for seven or eight months
and had cost the lives of 3,000 of their men.®
Ahmad ibn Sa'Id proved himself as capable a diplomat as he was
a fighter, and managed to ingratiate himself with TaqiKhan to such an
extent that he not only got himself confirmed as Governor of Sohar,
but also secured his appointment as Governor of Barka.®
It is stated in the Gombroon Diary that 3,500 recruits were to be
sent over to Arabia to replace those men who had fallen at Sohar
and elsewhere, “the King ha\dng ordered that when they were
Masters of the Sea Shore to march inland, and it is supposed his Designs
1 Guillain, vol. i, p. 538. Some uncertainty exists as to whether Sultan ibn Murshid
was kiUed in this manner or whether he perished when trving to force his wav into
Sohar through the Iranian lines. Salll ibn Razlq's stoiy of this fighting seems' to be
most inaccurate.
2 Guillain, vol. i, p, 538 ; Salil ibn Raziq (Badger), p. 1.50.
® Gombroon Giat}-, 2nd/13th April, 1713.
* Ibid., 8th/19th 3Iay. It was probably this rumour which misled Otter and caused
him to state that the Iranians had had to raise the siege (vol. ii, p. 181). It is to be
noted that Otter left Basra on bis return journey to France on the 9th /20th May 1743
‘ Gombroon Diary, 21st July/lst August. '
' Niebuhr’s Buchrdhnn^, p. 301 ; Guillain, vol. i, p. 538. Salll ibn Raziq (Badger)
pp. 149 and 150. *
NADIR shah’s CAWAIGNS IN ‘OMAN
169
are to conquer the whole Country, but while he is doing this he is
destroying his Own, and Nothing hut Misery, T}Tanny, and Oppression
are to be seen or heard in these Parts, the People being daily tax'd
that before Time is given for collecting one Another is laid on.” ^
In the meantime war had broken out at last between Iran and
Turkey. Nevertheless, Nadir did not abandon his designs on ‘Oman,
and, except for the withdrawal of some of his fleet from Sohar in
August, 1743,^ he took no steps to reduce his forces in Arabia in order
to strengthen his army in Mesopotamia.
For some time past there had been serious friction in ‘Oman between
Taqi Khan and Kalb ‘All Khan, and each sent to Nadir accusations
against the other. The consequence was that Nadir issued orders for
the recall of Kalb ‘All and appointed Muhammad Husain Khan
Qirqlu as Sardar in his place.® It appears that the Shah soon
afterwards also ordered the Beglarbegi to return.^
The new Sardar reached Gombroon in October and left at once for
SoMr via Julfar. Early in December, Taqi Khan, followed a few days
later by Kalb ‘All, arrived at Gombroon. A month later the Beglarbegi
revolted, and left the coast for Shiraz ; before leaving, he ordered
Kalb ‘All to be strangled and had his body thrown down a well because
the ex-Sardar had refused to join in the revolt.®
It seems that Taqi Khan's ultimate success in capturing Muscat
served to turn his head.® He thought that, with his influence in Pars
and (as he wrongly imagined) the control of the fleet in his hands, he
could defy Nadir ; such a belief, as he found to his cost, was mistaken.
It is beyond the scope of this article to describe Taqi Khan's rebellion ;
all that is relevant in connection with it is that it effectively diverted
Nadir’s attention from ‘Oman for a time. Then, when the revolt was
* Gombroon Diary, 21st July/lst August.
2 Gombroon Diary, 24th August/4th September, 1743. Nadir’s object was to
strengthen his fleet at Bushire which wa.s then being organized for an attack upon
Basra in conjunction with land forces. The vessels from .‘^ohar must have arrived too
late to participate in the united operations, in which the Iranian navy seems to have
played but little part.
^ Autobiography of Mirza Muhammad ShirazI, p. 16. (Professor SaTd Naficy,
of Tehran, possesses a M.S. of this autobiography, of which he has \er\- kindly sent me
a typew'ritten copy.) See also the Ta rthh-i-yadiri, p. 240.
* See the Td'rtlh-i-yadiri. p. 249. There are numerous entries in the Gombroon
Diary between September, 1743. and January, 1744, regarding the suspicious conduct
of Taqi Khan ; his recall was reported in Gombroon early in September.
5 Gombroon Diary, 30th December, 174,3/lOth January, 1744. Sec also Fnrsnama,
p. 194.
* Niebuhr’s Bescfireibung, p. 301.
170
LAURENCE LOCKHART —
suppressed, Nadir became engrossed with the Turkish war and had
no energy or resources to spare for side ventures. In consequence,
the Iranian garrisons in ‘Oman, though not withdrawn, were not
reinforced.
The astute Ahmad ibn Sa‘id was not slow to profit by these
developments.
One of the conditions of Ahmad's settlement with TaqI Khan
was that the former should pay tribute regularly to the Iranian
authorities at Muscat. After the departure of TaqI Khan, however,
Ahmad failed to make the payment of tribute on the appointed date,
alleging that he had no means of sending the money to Muscat-
The consequence was that the Iranian commanders at Muscat ran
short of funds and were unable to pay their troops, manv of whom
deserted.^ Ahmad's next step was to invite the Iranian commanders
to Barka, on the pretext of conferring with them as to the manner
in which the tribute could be paid. The Iranian commanders, accom-
panied by only a small escort, went to Barka, where Ahmad seized
them ; he then proceeded to Muscat and called upon the Iranian
garrison to surrender, offering them money if they jfielded of their own
free will, with the alternative of imprisonment if they did not. The
Iranians, bereft of their leaders, short of provisions and money, and
having no hope of succour from Iran, freely surrendered for the most
part. Ahmad, it is said, put some of the Iranians to death, but allowed
the others to return to their own country.^
In this way Ahmad became master of the coast from Sohar to
Muscat ; before long he extended his sway over the whole of ’Oman
with the exception of Julfar, which the Iranians managed to retain
for some years.®
Having delivered his country from the invaders, Ahmad claimed
his reward ; on summoning the chief Qadi of ‘Oman, he had no difficulty
in persuading the latter to arrange for his election to the Imamate.
This election is said to have taken place in the latter part of 1744 . «
So far as Iran was concerned the ‘Oman campaigns had proved a
costly failure ; many men were killed in action, but many more died
^ Xiebuhr's Be-schreibung, p. 302.
2 Niebuhr's Ik.orhrethvvtj, p. .303. See also Guliain, vo!. i, p, 540, and Salil ibn
Raziq (Badger, pp. 153 and 1.54). Salll's aecoiint of Ahmad's treachery to the Iranians
and his subsequent massacre of them seems much exageerated.
® Entries in the Gombroon Diary show that, as late as 1748, ships carrying men and
provisions were being sent over to .Julfar from time to time.
* Guillain, vol. i, p. .542 ; Otto Kersten’s “ Tabellari.sch Uebersieht der Gc.schiehte
Ostafrikas ” (in Baron von der Decken’s Seisen in Ost-Afrika, Leipzig, 1879), p. 18.
NADIR shah’s campaigns IN ‘OMAN
171
of starvation and disease ; it is probable that the total casualties
amounted to some 20,000 men. This heavy sacrifice brought no
advantage whatever to Iran ; for ‘Oman, too, the campaigns occasioned
much loss and suffering, but they led in the end to the supersession
of the decadent Ya'riba dynasty by that of the Al-bu Sa‘ids, which
rules Muscat to the present day.*^
The reason for Nadir's failure in ‘Oman was his giving the supreme
command to TaqI Khan, who was corrupt as an administrator and
inefficient as a militarv leader. Had he appointed instead some com-
mander of proved military capacity and integrity, like Tahmasp
Khan Jalayir, there is no doubt that ‘Oman would have been quickly
conquered and at a far lower cost in human lives.
^ R. Said-Ruete, in hia article The Al-bu-Said Dynasty in Arabia and East
Africa ’’ (Journal of the Central Asian Society, vol. xvi, part iv. p. 419), points out that
as the Ya'riba Dynasty attained power by expelling the Portuguese, so the .Al-bii-
Sa'id attained it by driving out the Iranians.
The Arabic Theatre in Egypt
By Nevill Barbour
Part I
The Beginnings ^
ri^HE establishment of the theatre in Eg3rpt, Hke many other
“*■ Western innovations, was aided by the initiative of the Khedive
Ismail. The Suez Canal was completed in the year 1869 ; to celebrate
its opening the Khedive decided to form the Ezbekia Gardens and to
build the Opera House which still exists beside them. This building,
in whose fabric wood occupies a dangerously large part, was erected
in a great hurry and inaugurated in November, 1869. The opera
Aida which Verdi was composing for the occasion in return for a fee
of 150,000 francs, on a hbretto originally arranged by Mariette Pasha,®
was not completed in time for the opening, and a performance of
Rigoletto was given instead.
About the same date, a second theatre, the Comedia, was built
within the area of the Gardens. These two theatres were, of course,
court-theatres, supported by funds supphed by the Khedive and a
few notables, and in no sense dependent on popular support. One
European company alone is said to have received from the Khedive,
in salaries and gifts, no less than 120,000 guineas for one winter's
season.
The first Arabic productions of which we have a record were pieces
composed by the Egyptian Jew Ya'qub ibn Rafa’il Sanu‘, commonly
knoivTi as ash-Shaikh Sanu Abu Naddara. Abu Naddara was a man
of distinguished appearance who, having studied in Italy, taught
European languages in an Egyptian school and made himself useful
1 (a) Ta’rikh Adab al-Lujhat al-‘Arabijya. Jurjl Zaidan. 2nd edition, Cairo,
1924. Part iv, pp. 152—7.
(b) Ta’rikh Migr fl ‘ahd Isma il. Ilyas Ayyub.
(c) Hayatuna at Tamthilij'ya. Muhammad Taimur. (Part ii — Mu’allafat Muham-
mad Taimur. Cairo, 1922.)
(d) Ihsan. Ahmad Zaki Abu Shadi. Cairo, 1927 (Appendices).
(e) Arzatu Libnan. Bairut, 1869. (Contains three plays of Marun an-Naqqadi
and 28 pp. introduction.)
(/) Ta’ril* al-Masrah al Mi?ri. Taufiq Habib. (MS. 160 pp. and appendices.)
{g) As-Sitar (Weekly Paper), Nos. 5-16 (31st October, 1927-16th January, 1928),
containing articles by Taufiq Habib about early actors and authors.
“ Egyptian Gazette, 27th July, 19.33, p. 2.
174
XE\^LL BARBOUR —
in various ways as an intermediary between Egypt and Europe.
In his spare time he composed plays in Italian and Arabic, some of
which were produced in the Opera about 1870 and gave great pleasure
to the Khedive, who named their author the Egj’ptian Moliere. One
of his Arabic pieces, entitled Miihjlr Misr v:a ma yuqdslh, was printed
in Bairut in 1912 ^ ; it is pleasantly written in rhjmied prose in an
educated form of Eg\'ptian colloquial, and treats of the actors in
Cairo and of their mode of hfe. The author is described on the title-
page as “ the King's poet, founder of the Arabic theatre in the Valley
of the Nile ”, Later he displeased the Khedive and was banished from
Egypt for some years, with the consent of the Italian Government of
whom he was a protected subject. He died in 1912.
AVe next hear of theatrical performances in Arabic in 1876. The
cultural impulse was again Itahan, though this time it reached Egypt
through the mediation of Syria. As early as 1848 Alarun an-Naqqa^,
a Maronite Christian from Sidon, who had spent a couple of years in
Italy, wrote a play called al-Bakhll, which he produced with success
in his house at Bairut before an audience of local notabilities. This
encouraged him to build a theatre where he produced other pieces of
his own composition, notably Abii Hasan al-Miighaffil (also called
Harun ar-Rashad) and al-Hasud. These were musical farces, partly
in classical Arabic, partly in Srnan dialect, and have been collected
and printed in Arzatu Libnan. Marun an-Naqqasb, whose character
commanded universal re.spect, died in 1855, at the age of 38 ; his
theatre, in accordance with instructions in his will, was transformed
into a church. This acti\'ity preceded by a quarter of a century the
establishment of the Jesuit and American Universities at Bairut,
though the Jesuits had an establishment at the village of Ghazir,
some 20 miles to the north of Bairut, from 1843. In 1876 Salim an-
Naqqa^, a nephew of Marun, with his friend Adlb Ishaq and a company
of actors including Yusuf al-j^aj-yat, famous it is said for his skill
in female parts, visited Alexandria and gave a season at the Theatre
Zizinia. The pieces presented were translations of European plays,
including Andromache, Charlemagne, Phklre, Horace and Zenobia.
adapted by an-Naqqash and Adlb Ishaq and enlivened with songs.
The season not proving a success, the two authors abandoned the
theatre for journalism. In 1878 al-Khayyat. after touring Zagazig
and Damietta, transferred his company to Cairo, where he was well
received by the Khedive and given permission to perform in the Opera.
1 iluhjir ilisr lua ma yuga-iih. Bairut, 1912, pp. 40.
THE ARABIC THEATRE IN EGYPT
175
Unfortunately he chose for performance a piece called az-Zalum
(“ The Tyrant ”) ; the Khedive took offence at allusions and situations
which he thought that his critics might apply to himself, closed the
Opera, and banished al-Khayyat from the country.
The same year, 1878, was noteworthy for the production in Bairiit
of an Arabic verse drama entitled AJ-Muru’ a ival-Wafd’, by Khalil
al-YazijI.i This piece, containing about 1,700 hues, is in three acts
preceded by a 40-line verse prologue (^utba) which classifies plays as
historical and imaginative and describes the qualities necessary in them,
such as naturalness of exits and entrances. The theme is the conversion
of Nu'man, King of al-Hira, to Christianity as the result of the nobility
of a Christian Arab who duly surrenders himself to be killed, after
being released on parole, on account of an unfortunate oath which the
king had sworn like Jephtha in the Bible. A subplot deals with the
love of the king’s daughter for the hostage who had made himself
responsible for the other man’s return. The play, which ends happily,
is naive, but its declamations in honour of virtue and duty are
agreeable.
In 1882 the new Khedive, Taufiq, allowed the Opera to be used
by a troupe organized by the S}Tian Sulaiman al-Qurdahi out of the
remains of the troupe of al-Khayrat with the addition of new members,
amongst whom was ShailA Salama Hijazi. Al-Qurdahl introduced
women, in the first place his wife, in the female parts, hitherto
exclusively taken by men, and added a female singer called Laila to
the company. The season of 1882 was followed by the rising of ‘Arabl
Pasha ; another season was given in the Opera in 1884 during which
large audiences came to hear the singing of Salama Hijazi ; the
company later toured the provinces and for some time gave
performances in a wooden theatre at Cairo. The pieces played were
those of their predecessors, with some new adaptations such as Othello
and Telenmque. It was in the time of this company that the work
of the well-known translator Najib al-Haddad began.
An interesting feature of this period was the effort of the EgN-ptian
patriot and keen Mushm ‘Abdullah Nadlm to use dramatic writing
as a means of educating the youth of Egj-pt. For this purpose he com-
posed two pieces, al-Watan and al-Arab, which were performed in
a school at Alexandria. The Khedive marked his approval by a
present of a hundred guineas to the school concerned. These pieces
* Al Muru’a wal Wafa’ au al-Farah ba‘d ad piq. Khalil al A'aziji. Bairut, 1884.
al-Matba‘at al-Adabiyya.
176
NEVIL.L BARBOUR —
are written in straightforward colloquial Egyptian prose with occasional
odes in the granamatical language. Though of httle artistic value
they are notable for their patriotic sentiment and practical interest
in the life of the people.^
The next troupe of note was that of Abu IGialil al-Qabbani, a
native of Damascus who had established a theatre in that city with
some success, until it was forbidden as the result of the complaints
of notables of the city to the authorities in Istanbul.^ AI Qabbanl
then came to Cairo and set up in a wooden theatre near al ‘Ataba
al-Khadra’.^ He introduced into his productions further elements
of operette and ballet.
About 1886 was formed the company of another Damascene,
Iskandar Farah, originally one of al-Qabbani’s actors, which included
Shaikh Salama Hijazi and was supplied w'th plays by the translators
Najib and Amin al-Haddad, Tanyus ‘Abdu, and Ilyas Fayyad. This
company dominated the Egyptian theatre for eighteen years until
Shaikh Salama broke away and formed a company of his own in 1904-5,
building the theatre called the Dar at-Tamthil al-‘ArabI near the
Wajh al-Birka street. During the long supremacy of Iskandar Farah
actors and pubhc made slow progress ; e.xisting pieces and traditions
were taken over, singing played the chief part, and Farah himself
regarded the theatre primarily as a commercial institution. The
highest salary paid in his company was thirty guineas a month ;
the fee to translator or author varied from twenty to sixty guineas.'*
The best pieces of this period were, in the opinion of Taufiq Habib,
al-Burj al-Hail of Farah Antun and Ghaniyat al-Aridalus of Khalil
Mirshaq. After the departure of Hijazi, Iskandar Farah formed a
fresh company which attempted to produce plays without music ;
this innovation found httle favoiur with the public. Costumes, scene;
and production of the period were not of a high order.
From this date it is easiest to consider the development of the
theatre in connection with the hves of the outstanding personahties
connected with it.
* Specimens in As-Sitar, No. 10.
= Objection was taken to the representation of the Khalifa Harun ar RaAid in
Abu Hasan al-Mughaffil by an-Naqqa^ in the character in which he is portrayed
in the Alf Laila ica-laita. To this day Harun ar-Rashid is the first theme to which
Arabic dramatists turn, e.g. al-'-Abfmsa in Cairo in 1931, a piece called Ar-Rashid
wal-Bardwika given in Tangier, Morocco, in 1929, and another with the same titl^y
Father Antun Rabat al-Yasu‘i (Bairut, 1924) and his portrayal as other than a
strictly orthodox and pious Muslim still arouses protest.
’ IVhere now stands the Cinema Olympia in ^ari* ‘Abd al-‘Aziz.
* MS. Taufiq Habib.
THE ARABIC THEATRE IX EGYPT
177
Saldma Hijazi was a poor man's son, born in Alexandria ;
having a good voice he found emplojnnent as a muezzin and in private
houses as a reciter of the Qur’an. From childhood he was attracted
to the theatre ; as a young man he visited European theatrical perform-
ances in his native city in the company of friends of similar tastes
amongst the educated S^uians of Alexandria. His first appearance in
Cairo was with al-Qurdahi ; then followed his eighteen years with
Iskandar Farah. He seems himself to have desired fame as an actor
rather than as a singer ; but in the judgment of the public he was
only one actor among many, while as a composer and singer he has
had no rival among Eastern musicians in modern times. The long
dependence on Iskandar Farah is held to have been prejudicial to
Shaili Salama. Though an uneducated man, he apparently had more
artistic sensibihty and more desire to serve artistic ends than his
employer. When in 1904-5 he opened a theatre of his own beside the
Ezbekia Gardens his productions marked a notable advance in scenery
and costume. After an interruption due to illness, ^ai]A Salama
returned to the stage ; in 1910 he performed in Tunis, in a theatre
in the Bey’s palace. From 1914 to 1916 he was associated with Jurj
Abyad. His death occurred in October, 1917. His best known
performances in romance were in t^uJiadd’ al-Qhardm {Romeo and
Juliet). as-Sayykl, and Raldh ad-Dln; in tragedy, in Hamlet', in
comedy, in Sidq al-IlJa and Uns al-Jalls. His most successful
compositions were said to be the operettes 'Aida and ‘Izat al-Mululc.
The brothers ‘U kasha. These were three brothers, ‘Abdullah,
‘ Abdul-Hamid, and Zakl. The period of their greatest actiHty extended
from 1910 to 1925. They were actor-managers and were the promoters
of the theatrical society which in 1924, with the aid of Tal‘at Bey
(now Pasha) Harb, in connection with the Societe Misr, built the
existing handsome theatre in the Ezbekia Gardens, of which Zaki
‘Ukasha is now lessee. From the artistic point of ^^ew the theatrical
activity of the three brothers has been harshly judged and it has been
said that they neither knew nor wished to know an}’thing of art
or artistic principles.
‘Abdurrahman Ru.^dl is an actor of personality, of the same period,
who has enjoyed considerable success in romantic roles. He has at
various times formed companies of his own and still occasionally
appears in Cairo and the provinces.
‘Aziz ‘Id is a veteran of the Eg^-ptian stage and is at present,
1932-3, producer and artistic director in the company of his wife,
VOL. viii. part 1. 12
178
KE\^LL BARBOUR —
Fatima Eu^dl. S\T:ian by birth, his connection with the stage goes
back some thirty years, in which time he has been concerned in many
theatrical ventures. It is said that his fame was first made in a piece
in which he represented an Omda from the provinces, drinking in a
bar in the company of the European women who frequented it. He
is considered to have originated on the Egj-ptian stage the form of
buffoonery known as Franco- Arab re\'ue
which is now successfully exploited by ar-Eihani and ‘All al-Kassar.
There is no doubt that ‘Aziz Td possesses real artistic sensibility and
in comic character parts is an actor of the first class. It was he who
was responsible, in collaboration with the well-known musician and
singer, the late Sayyid Darwish, for the successful production in 1920
of Muhammad Taimur's operette The Ten of Diatnonds.
Jurj Ahyad is a SjTian who entered the service of the Egyptian
Government and became station-master at Sidi Gabir, Alexandria.
In his spare time he took part in theatrical performances ; his acting
having on one occasion attracted the attention of the Khedive
‘Abbas II, the latter sent him to Paris to study dramatic art. This
he did under the well-known actor Sylvain.
Great expectations were aroused on his return in the year 1910,
and were largely fulfilled when, having collected a company, he
produced Oedipus, Louis IX, Othello, and other pieces. The end of
the season, however, was not so successful as the beginning. The
subsequent history of Jurj Abyad’s theatrical ventures was largely
a repetition of the same cycle of events.^ He did not possess the deter-
mination necessary to carry a company through moments of dis-
couragement, nor originality enough to create new parts. Nevertheless
in those parts which were modelled on the performances of Sylvain
he achieved something approaching perfection ; and the standard of
acting and production which he set marked an epoch in Egyptian
theatrical history. During the Great War and the consequent absence
of visiting European companies, Jurj Abyad gave a successful series
of performances in French in the Opera House at Cairo. In 1920 he
toured North Africa. In Algeria he reported Arabic culture to be dying
and the Arabic language hardly understood. In Tunis, on the other
hand, he was heartily welcomed ; as also in Tripoli, where he was
agreeably surprised by the number of Italian officers who attended
^ Thus in^l
newspaper, lith August, IMS.
2 Hayatuna at TamUiIliyya, pp. 131-142.
THE ARABIC THEATRE IN EGYPT
179
his performances in spite of their being given in a language of which
they could understand very littled
Jam ‘latu Ansar at-Tamthll. In March, 1914, a society with this
title was formed to further the interests of “an independent and
national theatre The founder was Muhammad ‘Abdurrahlm, who
had studied in England ; this seems to be almost the only example of
English influence on the Eg^-ptian theatre, practically all the actors
having learnt their technique in France or Italy, as a glance at the list
of theatrical terms at the end of the third part of this article will
show. In spite of the War and the founder’s early death, the society
has performed a certain amount of useful work and in a modified form
stiU exists to-day.
Muhammad Taimiir, though not himself an actor, exerted a most
beneficent influence on the Egyptian theatre from his return from
Berlin and Paris in 1914 to his untimely death in 1921. Equally gifted
as plajTvright and critic, endowed with a charming personality and
a high social position, his death when still less than thirty years old
was a quite irreparable loss to the Egyptian stage, for whose emancipa-
tion from commercialism he was an ardent champion.
In concluding this summary of the early history of the theatre in
Egypt, we cannot do better than utilize Muhammad Taimur's classifica-
tion 2 and divide the first fifty years of the Egj’ptian theatre into four
periods. During the first, from 1870 to 1880, the theatre depended
for its appeal upon its novelty ; in the second, from 1880 to 1904
(Iskandar Farah), upon the singing which, strictly speaking, was an
irrelevancy ; in the third, from 1904 to 1910. upon the singing aided
by competent scenery and good costumes ; while in the fourth period,
which begins with Jurj Abyad, the Egj-ptian theatre, though still
in its infancy, may claim to be sufficiently advanced to be judged by
the artistic standards of the countries in which the drama is an old-
established institution.
Part II
Present Conditions (1932-1933)
The theatres at present exi.sting in Cairo are the following : —
Opera House Midan al-Opera.
Ezbekia Theatre Ezbekia Gardens.
Dar at-Tamthil al-‘Arabi ® ,, ,,
^ Newspaper cutting attached to MS, Tauflq Habib.
- Hayatuna at Tani^iliyya, pp. 22-6.
^ Now used, occasionally only, by inferior revue companies.
180
XEVILL BARBOUR —
Eamses Theatre ^ Shari' Tmad ad-Din.
Ar-Rihani Theatre ,,
Printania Theatre ^ „ „
Majestic Theatre ^ „
Bosphore Theatre * Midan Bab al-HadId.
There are in addition a number of Salas or Music-halls, the best known
being that of Badi'a Masabnl, the wife of ar-EihanI, also in Shari'
‘Imad ad-Din.
From June to September the theatrical quarter of Cairo is deserted,
the companies moving either to Alexandria or to open-air theatres
at popular resorts in the suburbs of Cairo. Outside Cairo there is no
regularly established theatrical company. Alexandria and the
provincial towns are dependent entirelv on vdsits of touring companies
at irregular intervals and on local amateur dramatic societies. The
municipality of Damanhur has recently built a handsome municipal
theatre.
Companies. Two companies have given regular annual seasons of
serious drama in Cairo for the last seven years ; these are the Eamses
Company of Yusuf Wahbl and that of Madame Fatima Eu^dl.
Both these companies are run on the actor-manager and repertory
system.
Yusuf Wahbi is the son of a well-to-do Pasha of Turkish origin
though the family has long been settled in Egj’pt. Attracted by the
theatre, he spent some years in Italy, where he acquired a good know-
ledge of theatrical art. Returning to Egj’pt, he founded the Ramses
Theatre in 1923 in conjunction with ‘Aziz 'Id ; since that date the
Ramses Company has been recognized as the leading Eg^-ptian companv,
distinguished by the general high standard of the actors and by the
excellence of their Arabic diction. A'u.suf Wahbl himself is a capable
actor who knows how to dominate the .stage ; he is, however, criticized
for a tendency to melodramatic over-emphasis. There are also heard
about him the complaint.s which are made about actor-managers in
all countries ; it is, indeed, obvious that the difficulties of such a
position are likely to be considerable in a country where there is onlv
one alternative company to which an actor can transfer. Yusuf M'ahbl
is a man of American energy. He is a remarkably good linguist and is
’ Built by Yusuf Wahbl in 192.3.
= Of Greek ownership and used often by visiting Greek companies, but occasion-
ally also by Arabic companies fe.g. in the winter of 1932-3 by Fatima Rushdi).
^ Used by ‘All al-Kassar.
^ Under reconstruction.
THE ARABIC THEATRE IX EGYPT
181
capable of acting in English and Italian as well as in Arabic. The
leading lady of the Eamses Company is Amina Eizq, a young and
attractive actress who excels in the role of the unsophisticated heroine ;
her declamation of classical Arabic is very pleasing. Amongst the
other members of the company, Ahmad ‘Allam has distinguished him-
self by his performance of Qais in ShauqI’s Majnun Laila. ‘Umar
Wasfi, a competent comedian, is a veteran of the Egyptian stage ;
Istafan Eusti, of Italian origin, in addition to acting, adapts European
pieces for the Egj^ptian stage.
The second company is that of Madame Fatima EushdI. This
actress was originally a member of the Eamses Company, which
she left after three years, with her husband, ‘Aziz ‘Id, to form a
company of her own. Her training has been entirely in Eg}^)!, and
she does not speak any foreign language. She has not built herself
a theatre, but hires one of the existing Cairo theatres as required.
She is very youthful, unaffected in manner, gay and unsophisticated
— characteristics which won her the title “ Sadlqat at-Talba ". She
is at her best in Jluslim costume parts, such as Najaf, the beggar’s
daughter in A Night from the Thousand Nights ; but she is versatile
and has played to the satisfaction of the public such diverse roles as
Cleopatra in ^auql's Masra‘ Kliyilpatra, I’Aiglon in Eostand’s play
of that name, Mark Antony in a translation from Shakespeare, and
Tutu, the boy hero of a vaude%ulle. The production of the plays of
this company is undertaken by ‘Aziz ‘Id. The ensemble of Fatima
Eu^dl’s company is generally held to be inferior to that of the Eamses,
notably in their mastery of classical Arabic. It is, however, favourably
contrasted with the former for not pandering so much to the popular
taste for sensationahsm.
Besides these two permanent companies, a certain number of other
performances were given in Cairo during the season of 1932-3. Jurj
Abyad, with a specially collected company, gave a few performances
in the Opera House. These were, however, in French (Le Pere
Lebonnard, Le Reveil, L' Arlesienne) and were only noteworthy for the
distinction of Jurj Abyad’s own acting. ‘Abdurrahman Eu^dl with
a company gave a few performances in the Ezbekia Theatre, including
al-Badaiviyya, al-Maut al-Madanl (from the Italian) and al-Bu’r
al-M u ra khhh a sa .
A company led by Fikturiyya Musa gave performances in the Opera
of a piece called Sihdm (‘' Maxim’s ”) and of the old favourite Ghaniyat
al Aruialus.
182
NEVILL BABBOUR —
Cairo is also the headquarters of two musical re\Tie companies.
The best known of the two is that of Xajib ar-RihanI (" Kish-Kish
Bey This actor has been the most successful in his undertakings
of any that Egj’pt has known, and liis theatre is the only one regularly
visited by a considerable number of Europeans. Indeed, the success
of Franco-Arab revue is sometimes held to have debased the public
taste and injured the position of the legitimate drama. The second
company is that of 'Al! al-Kassar. This company caters for a popular
audience. Al-Kassar himself is an amusing comedian, but the rest of
the company are not up to his standard.
One other company, that of Madame Munira al-Mahdiyya. also
performed in the Ezbekia Theatre during the winter. This actress,
who can claim to have been the first Eg\q)tian actress, her predecessors
being all SjTians, started her career as a singer ; the theatrical pieces
in which she appeared later were mostly romantic plays in which she
generally took male roles, or farces with occasional songs. Sometimes
the theme was historical, as in Haldh ad-Dln, an old favourite adapted
from Sir M'alter Scott’s Talisman. In 1929-1930 Munira al-Mahdiyya
co-operated with the singer, ‘Abd al 'Wahhab, in producing an
opera. Tosca was chosen, the plot and costumes of Puccini's opera
being used, while the music was Egj'ptian. The piece was very short,
lasting about an hour and a half. The experiment was interesting, but
the music did not appeal to the public and the production was not a
success. Another piece, Antony and Cleopatra, had a better reception,
some of the music being by the popular composer, ShailA Darwlsh.^
Salaries. The highest salaries paid hitherto were in the revue
company of Xajib ar-Rlhanl. amounting to some 100 guineas monthlv.
A leading actor in the e.xisting dramatic companies is fortunate if
he receives 30 to 40 guineas a month, or a leading lady 20 to 30 guineas.
The Government. The attitude of the Government towards the
theatre is in general benevolent. As regards the censorship of plays,
the theatre comes under the Ministry of the Interior, to whom copies
of each play have to be submitted for sanction before production.
The censorship is chiefly concerned with avoiding offences to morals
or to Egyptian or foreign susceptibilities. Thus it is apparently not
permissible to represent on the stage King Louis IX of France in
* .t fiirio.sity <if Egyptian dramatic literature i.-. the existence of six ojK‘ra lilirettos
composed by the versatile and indefatigable Dr. Ahmad Zaki Abii Sliadi. These are
entitled “ Arda^ir '' Az-Zabba' '' (Zenobia), " Ihsan ”, " .tkhnatfin ”, '■ Hint as-
Sahra’ ”, “ Al-Aliha ”. They have been printed, but have not yet found a producer.
THE ARABIC THEATRE IN EGYPT
183
captivity after his defeat at Mansura. In other respects the theatre
comes under the section of Fine Arts of the Ministry of Education.
In 1924, as the result of a competition, Zaki Efendi Tulaimat,
secretary in the Zoological Gardens and an amateur of the theatre,
was sent to Paris for four years’ study at the Odeon. After his return
to Egypt and on his recommendation, an Academy of Dramatic
Art [Mahad fann at-TamtJnl) was estabhshed by the Ministry of
Education. This institution was planned on ambitious hues ; in
addition to courses on declamation, technique of the theatre, decora-
tion, lighting and make-up, there were classes in dancing, physical
culture, and the French language ; while Dr. Taha Husain lectured
on the history of the drama and Dr. Ahmad Daif on Arabic literature.
The Academy, under the management of Zaki Tulaimat, opened in
1930 with forty pupils of both sexes ; after one year a new Minister of
Education, Hilmi ‘Isa Pasha, decided to close it. Two reasons were
given. The first, that the mixing of the sexes in an institution of this
sort was a danger to morahty and contrary to the precepts of Islam.
In particular the classes in Eurythmics, participated in by young
men and young women in the same room, roused opposition.^ The
second, that acting is a matter of natural talent, aided by practice,
and that a Government institution of this sort would not achieve the
purpose for which it was designed. The ensuing controversy filled
many columns of the newspapers in midsummer 1931 ^ ; and the
place of the defunct Academy was finally taken to some extent by
the “ Lecture Hall ” {Qaat al-Muhadarat), a room in the Ibrahimiyya
School, in which lectures are given in the afternoon for two hours,
three times a week, on theatrical subjects. These lectures, which are
given by Zaki Tulaimat, Jurj Abyad, and others, are open to the pubhc
and are attended by some thirty pupils of various ages and stations in
life. Girls may attend, but at different hours to the male students.^
^ “ The religion of Islam does not permit Muslim women to dance in the presence
of men not of their family under any circumstances whatever. ... It is a general
principle of Islam that the prevention of corrupting influences precedes the bringing
of improvement.” Shaikh of al Azhar in the Ahram newspaper, 21st August, 1931.
2 e.g. Al Ahrdm, 21st August, 1931 (interview with Shaikh of al Azhar). Al Ahram,
22nd August, 1931 (interview with Minister of Education). Wadi an Nil, 28th July,
1931 (suggestion to replace the actors and more particularly the actresses by
marionettes).
® A Committee of foreign professors of literature was formed in 1930 to choose
twelve plays suitable for translation into Arabic. The selected plays, when translated,
were to be produced by Jurj Abyad. After the Committee had held several meetings
the project seems to have been abandoned.
184
NEVILL BABBOCR —
In 1925 and again in 1932 a competition for playwrights was held
under the auspices of the Ministry. In that of 1932, no less than 143
plays were submitted.^ A first prize of 100 guineas was awarded to
a play of modern Egyptian life, entitled Samira, by Muhammad
Rashad Hafiz, and second prizes of seventy-five guineas each to
al-Awcitif (modern Eg}y)tian) by Muhammad j^urshid, al-Hudi
(Abbasid historical) by ‘Abdullah ‘Afifi, Ibnat ash-^ams (Ancient
Egyptian) of Eransls Sliiftashl, and Ahmus al-Awival au Tard ar-
Ru‘dt (the expulsion of the Hyksos) by ‘Adil al-Gbadban. These plays
were all in the classical language, the colloquial being apparently
excluded by the terms of the competition. The committee was strongly
criticized as having judged entirely on linguistic and moral grounds
and as having no practical experience of the theatre. This criticism
was supported by the refusal of the theatrical managers to produce the
winning piece. The author appealed to the Minister, who lent him
the Opera House, and advanced him 40 guineas from the public funds,
the author being debarred by the regulations from receiving the prize
money until the piece had been produced. The public, however,
in spite of clever producing by Zaki Tulaimat, confirmed the opinion
of the managers, only one scene appearing to possess any appreciable
dramatic value. Nevertheless it cannot be doubted that the competition
promoted an interest in play-writing.
The Ministry also seeks to help the theatre by annual grants to
the theatrical managers and to individual actors. These grants
naturally lead to a good deal of ill-feeling, especially as the distribution
varies from year to year and is accompanied by unwelcome criticism,
while the total sum distributed only amounted in 1933 to 1,048 guineas
as opposed to the 8,000 guineas allotted to visiting foreign companies
whose performances are attended almost exclusively by foreigners.^
Of the 1,048 guineas, 200 went to the management and 140 to the
members of the Ramses Company ; 200 to the management and 90
to the members of Fatima Ru^di’s Company — individual actors
receiving 20 or 10 guineas each. The remainder was divided amongst
four of the other companies above-mentioned,® and amongst three
dramatic societies.
Tours. The tours made by the different companies in various
* A] Ahram, 20th January, 1933. A? Sahah, 27th January, 1933, p. 20.
* A? Sabah, 5th May, 1933, p. 24.
’ Rihani’s Company were not given a share as, being abroad, they were held
to have escaped the observation of the Government.
THE ARABIC THEATRE IX EGYPT
185
parts of the Arabic-speaking world are an interesting feature of
Egj'ptian theatrical hfe. The Ramses Company is well known in
Palestine and Sj ria, and in the year 1928 Yusuf Wahbl took them
across the Atlantic for the benefit of the thousands of Syrians living
in the Argentine Repubhc, giving a series of performances in Arabic
in Buenos Aires. The company appealed to local taste and the Syrians
were occasionally crowded out of the theatre by the wealthy Spanish-
Argentinian aristocracy. Madame Fatima Ru^dl has taken her
company to Palestine, SjTia, and Iraq, and in the summer of 1 932 her
company made a very successful tour through Tunisia, Algeria, and
Morocco. In the winter of 1932-3 Najib ar Rihanl and his company
followed the same route through North Africa. The journey was not
an entire success ; debts were contracted by an impressario in Tunis,
and there was a misunderstanding with the inhabitants, afterwards
rectified. In Algeria the conduct of some of the chorus girls led to
their being sent back to Egjqit. Moreover the greater orthodoxy of
the local Mushms compelled, it is said, the company to observe
Ramadan in Algeria with a strictness that they did not observe at
home.^ In Marrakesh, according to the correspondent of as Sabah,
they were received by the Governor, at-Tihami al-Jilawi, at his house.
After laudatory odes delivered in the host's honour, such as —
.sIjU;
the Governor presented 10,000 francs to ar Rihani himself, and
envelopes containing 1,000 francs to each member of the company.
Criticism and Theatrical Neu's. There is a scarcity of writers who
have had sufficient experience of the European theatre to form a sound
dramatic judgment. Thus the criticism which is frequently found in
the daily and weekly papers consists largely of the description of
plots, criticism of the language and personahties.
The theatres only advertise in the daily papers internu’ttently
and incompletely ; the public rely on posters displayed in various
parts of the town. There is, however, one weekly paper, as Sabah,
of some eighty pages, which occupies itself primarily with the theatre.
In it can be found announcements of all important theatrical events,
summaries of plots of new plays, together with a variety of informa-
^ A? Sabah, 31st March, 1933.
186
XEVILL BARBOl'R —
tion concerning the companies, actors, singers, and dancers. It is also
supplied with theatrical information by correspondents in North
Africa, Syria, and Iraq.
The Public. The Egyptian theatrical pulilic is vmry limited. The
religious leaders disapprove of the theatre because of the mingling
of the sexes upon the stage. The aristocracy, for the most part
educated in European schools, are apt to take an exaggeratedly
“ European attitude towards the native theatre, and for that reason
to neglect it. For the populace and the fellah the theatre is still a
Frankish innovation. The theatrical public then must be provided
chiefly out of the Government emplovees and small bourgeoisie whose
secondary education has given them a certain taste for things
European. It is only in Cairo that this public is sufficiently numerous
to support a whole-time theatre ; and this theatre must be run on the
repertory system, it being impossible to find an audience for the same
piece for more than a week consecutivelv. This public, like all
unsophisticated publics, likes melodrama. But it is composed of
Cairenes, a people famous for their facile gaietyq their incessant
idiomatic repartee, and their love of their native music. Thus the
Franco-Arab revue is sure of support. At the same time, they
are Arabic-speaking and have an Arabic culture which goes
very deep. They are, therefore, extraordinarily responsive to the
unaided appeal of fehcitous language on traditional Arabic lines.
Further, their intere.st in their history creates an opportunity for
historical plays dealing with the heroes of the Islamic past, while
the growth of nationalism in recent years is bringing with it an interest
in the Pharaonic period which gives an opening for another tvpe of
historical play. These characteristics of the Egy’^ptian public seem to
have determined the nature of the theatrical fare offered to it.
Language . — The spoken language of Egypt differs, of course, very
widely in grammar, pronunciation, and vocabulary from the modern
form of the classical language which is employed in nearly all lectures
and newspapers, and in hterary and scientific works. The first question,
then, which faces the playwright is that of the language in which
his play is to be composed. His choice will probably depend upon
the subject of his play. Serious hi.storical pieces are most naturallv
written in a language as near the high classical style as the author
can command. Plays of modern life tend to be in the vernacular,
as are most farces and revues, though the latter make use of the
possibilities of comic effect by the juxtaposition of the two languages.
THE ARABIC THEATRE IX EGA^PT
187
The late Muhammad Taimur, far the most vital of Egyptian play-
wrights up to the present time, composed the first draft of his first
play ‘Usfur Jil Qafas, in the classical language, and then rewrote it
in the vernacular, which he employed for ail his subsequent pieces.
On the other hand, vernacular plays are occasionally translated into
grammatical Arabic for production outside Egypt,^
^ e.g. “ al Wuhu^ ’’ by Mahmud Kamil. See al Jamra ”, 11th May, 1933, p. 4.
{To be concluded.)
Grammatical Categories in the New Hebrides
(A Preliminary Study of Linguistic Psychology in Malekula)
By A. Capell
Introductory
A LTHOUGH it is generally admitted that European and extra-
European grammatical categories do not absolutely coincide,
one very rarely sees the grammar of a native language worked out
on the clear recognition of this fact. The following paper is intended
to illustrate the differences in grammatical thought between English
and one language of the New Hebrides. It is not intended to be a full
study of that language, but only to indicate what seems to be the
mental outlook of the people who speak it. The language in question
is that of Achin, one of a group of small islands off the north-eastern
coast of Malekula in the northern New Hebrides. The material which
has formed the basis of this study I owe to the kindness of Pastor
A. G. Stewart, of the Seventh Day Adventist ilission. Nothing
grammatical has yet been published concerning the language ; that
of the neighbouring island of Uripiv, ho%vever, resembles it closely,
and an outline grammar of that language may be seen in S. H. Eav's
Melanesian Island Languages (Cambridge University Press, 1926),
pp. 284-293. While, however, the Uripiv language does to some
extent exhibit the phenomena peculiar to this region, it does not
seem to do so as throughly as Achin, and the peculiarities themselves
are not brought out in Ray's sketch. Of the languages of the other
islands, Wala and Rano, I have no information, but Pastor Stewart
informs me that there is considerable resemblance between Achin
and IMatanavat, on the opposite mainland, and then a considerable
cleavage between that and the country of the Big Nambus, inland.
Southwards, the peculiar features of Achin and Uripiv do not reappear,
except to a minor extent in Ambrim.
In European languages we are accustomed to divide our grammars
into treatises on eight "parts of speech”. Sapir. in his Language,
clearly recognizes that such a division may not hold good for native
languages ; Je.spersen, however, in the Philosophg of Grammar prefers
to find them a fairly permanent feature of language in general. In
190
A. CAPEIX
Achin it may be convenient for practical purposes to retain the
division into eight parts of speech, but these eight must be rearranged
again in two larger groups, which ma}' be called ” object-words " and
" time-words Object-words correspond to our nouns, pronouns,
prepositions, and interjections ; time-words include not only verbs,
but also adjectives, adverbs, conjunctions, and numerals. It remains
true, however, that adjectives and adverbs still have one foot in each
division. The paper that follows will give examples of this arrange-
ment of speech and try to suggest a line of explanation for some of
the usages.
Spelling
The spelling here followed is not that of the Mission, but that of the
International Phonetic Association used in very broad transcript.
The only serious departure from their system is the use of d, which
indicates an e-sound that has arisen through Umlaut, i.e. the influence
of a following i or less often u. Thus, when the article na is prefixed
to the word im, house, the result is ndim, spelled by the missionaries
neim ; the word for “ live ”, common in Oceanic languages as nmuri,
here becomes mdur, spelled by the missionaries meur ; tay, weep,
with the transitive suffix becomes taysi, weep for, for an older tayisi.
With i the Umlaut is invariable ; with u there are exceptions — vami
and vdnit (venii) are both wTitten for village The Melanesian g,
spelled in the Mission -RTitings as h, is here written y ; while the velar
nasal («g in sing), spelled g by the missionaries, is here written y.
C indicates the palatal plosive heard differently by different observers
as ch or ts, and spelled in the translations as j. It is the sound
indicated by c in the I.P.A. Script.
1. Grammak
It is not part of my purpose to give a full grammar of the language,
which is sufficiently like Uripiv for the sketch of the latter given bv
Ray to be used in the reading of this paper. Achin is characterized bv
a considerable amount of abbre\dation, e.g. the verbal pronoun of
the 3rd per.son plural indefinite, Uripiv ara, is here ar. The pronouns,
as far as we shall need them, are :
Plur. 1 inch ikir ; 1 excl. kinm
2. kami
3. inir.
Sing. 1. ina
2. inik
3. ini
GRAMMATICAL CATEGORIES IX THE XEW HEBRIDES
191
Suffixed pronouns show two series, as in Uripiv. They are ;
1. With nouns and some prepositions, e.g.
Sing. 1. tuali, my brother Plur. 1 inch tuar ; excl. iuamam
2. tuom 2. t lie mi
3. tuan 3. tuar.
In the suffixes to prepositions the n of the 3rd person singular is lost.
With verbs, a replaces the i of the 1st person singular, mu the mi of
the 2nd person plural. Curiously enough, the pronouns suffixed to
the verb precede, instead of following, the ti'ansitive suffix ni : wire-
mu-ni, say it to you ; vise-nir-ni, teach them it.
Another important element of Achin speech is the ligative article y
of Indonesia, found also in various other parts of Melanesia. Thus
in the title of the Bikol (Phihppines) Bible, the Spanish adjective
santo, holy, receives the addition of the ligative article and we have
maya santoy kasuratan, the sacred writings. In Achin and Uripiv,
however, what was originally an article, intervening between adjective
and noun, becomes a conjunction by the use of the verbal particle
after it and before the adjective. Thus cinop ya mu cac, a bad man.
This ya is equivalent to the %vh- of the English relatives who, which,
where, so that the phrase just quoted might be rendered literally by
“ the man wh- he is bad ". Ray's Uripiv example, deluy ya mara
mdsi, is literally people wh- they are sick This explains three
uses of ya, all of which will be important for us in the following pages ;
1. With adjectives, as in the examples just quoted, and also ;
cinop ya mu nac, another man (lit. “ man wh- he is different ’').
2. With pronouns : ini ya mi te niel, he wh- he made the sun ;
nembe ini ya mi ara worpi re kiy ser mere Israel 1 Where he wh- they
bore shall become king their people Israel, i.e. where is he that is
born to be king of Israel ? Kami ya kam ivorsuri e wale wele ko ro,
you wh- you believe only little indeed, i.e. you of little faith ; e pe ya
mar mac pin, inir ya mar pecelei tipis wele ok po revci pini, since thev
are dead who sought the young child to kill him.
3. With nouns : Tayar semam ya pi mtiur tacer, God our wh- he
shall live for ever ; 7iono ya nawinen e res : thing wh-its scent good,
i.e. a thing whose smell is sweet.
These notes do not fully explain the uses of ya, but give sufficient
indication to make clear its occurrences in the examples that will be
adduced in the next part of this paper.
In verbs, tense is indicated just as in Uripiv, except for abraded
192
A. CAPELL —
forms. Ray (op. cit.. pp. 288-9) finds in Uripiv^ three tense particles,
e, mi, and pi. Of these, c is a perfectly indefinite particle, serving
merely to mark the word it accompanies as a time-word. It is the same
as the e that occurs in most Polynesian and a good many Melanesian
languages. Mi is a particle with a variable vowel (examples of 7nii have
already occurred in this article), and is common throughout the northern
New Hebrides. Ray says : “It appears to indicate a more definite past
or present than e, and is very commonly used after the hgature ya.”
This appears to be right : e marks a word as a time-word ; mi draws
attention to an action at a fairly well defined time. A complete
past event may be shown by die (all) ; vadn (over, cf. ilota vela), or
ilia (of old, a common Oceanic root), concluding the unit of expression,
as in the following sentence : Mare e tapar vadn pitevi, the heaven
was opened to him. Pi is definitely future and also has a variable
vowel. Then there is a Jci which Ray does not give for Uripiv, and which
seems to vary between a future and a subjunctive. These three
particles coalesce with the pronouns as follows :
Mi
Pi
Ki
Sing.
1.
(no)
pe
ke
9 .
(ko)
po
ko
3.
mty mil
pi, pu
ki
Plur.
1 incl.
ram
rap
rak
1 excl.
mim
nap
nnk
2.
ham
kap, kop
3.
mar
par
kar
The nn form is slightly defective ; the e mav be replaced in the 3rd
plural by ar and or. In addition the emphatic particle nk may follow
the verbal pronoun, giving mar ok, without seeming to alter the idea
to any great extent. The.se particles are phoneticallv somewhat
different from those of Uripiv, but the general scheme of things is
much the same.
Achin Syntax
We are now in a position to cousider the special peculiarities of
Achin syntax. A.s regard.s rules of accidence, strictly so called, it is
still possible for u.s to retain the European parts of speech, at any rate
for our own convenience. It is rather in the field of svntax that the
division into object-words and time-words becomes important, and
even in these, Achin is simply carrying to its logical conclusion
a tendency visible in more than one Melanesian language.
GRAMMATICAL CATEGORIES IX THE XEM' HEBRIDES
193
A. Object-words
By object-words are meant such as are not characterized bv
distinctions of tense. Under this heading are included our European
nouns, pronouns, and most prepositions. Thus ci»op, man, remains
under all conditions cinop. Similarly the stems of the pronouns do not
themselves change, though they mav' undergo moditication through
the verbal pronouns for person and tense. The cardinal pronouns
belong to the category of object-words, the verbal pronouns to that of
time-words. Some of the object-words, however, seem to be on the
border line between objects and times, or rather, let us say. the Achin
native seems to put under the heading of times ideas that we should
put under that of object. Thus, matur = sleep, nviturporpor = dream.
These are both verb and noun in English, while in other European
languages, if the two parts of speech are not identical, they are at
least related, as for example French rece and rh'er, German Trnum
and triiumen. In Achin, if it is needed to say in a dream then the
expression is reya mi m-aturporpor for the 3rd singular, reya mar
maturporpor for the 3rd plural, and so on. One would naturally
translate such an expression as "while he dreamed", "while they
dreamed ", but then comes the difficulty that reya certainly means
“■ in ", e.g. reya Rano, in Rano. Clearly, what is for us a noun, “ dream,"
is regarded by the Achin people as a time-experience rather than
an object-experience. It is not that a ‘‘ dream is a thing which one
may come upon, but an experience that one has in a certain time.
And this is quite logical ; a dream is not an object that occupies
space.
Another example of a similar crossing of the barrier between
object and time categories is seen in the expression for " they that
dwell in great darkness ”, inir ya tnar lek reya mu ror e lep. lit. " they
wh- they dwell in it is dark it is great The e of e lep is the particle
of the indefinite tense, showing that the -word lep (Fiji leva, great) is
a time-word, but the time is continuous, and. in fact, not important
for the idea. The darkness may change ; it has the more definite
particle am, but as long as it is dark, the darkness is great, and so is
expressed by the indefinite particle e. It hardly seems po.ssible to
translate reya here as anything else than “ in ", yet it is followed by the
verb-form mu ror. it is dark. Psychologically, it is obvious that the
idea "it is dark " occupies the native mind, not the abstraction
" darkness ". It is not that a something called " darkness ” covers the
people for a certain time, but that for a certain time they have an
VUL. \-III PART 1 . 13
194
A. CAPELL —
experience of inability to see, because it is dark. In other words,
it is again a time-experience, not an object-experience, to the native
mind.
Another possihihty is that the entire verbal expression may be
treated as a sort of compound object-word. This is an extension of
the phenomenon just dealt M*ith. E.g. reya Icom viel le u'or, while you
are going with him, ht. in you-go-with-hini. I know of no other
language where there is this continual tendency to translate the
abstract into temporal forms. A gerund is conmion ev^n in other parts
of Melanesia, “ in your going with him ” ; in other places again the
abstract must he translated into the concrete. Yet, undoubtedly, to
the native mind the darkness (to revert to the second example) is
just the experience of a certain condition for a certain time, rather
than a thing that covers a certain area. It is a combination of our ideas
of “ in the dark ” and “ while it was dark
Most prepositions can be classed as object-words, because they are
invariable for tense — and indeed it is hard for us to see how a “ preposi-
tion ” can vary for tense. The preposition shows a persistent relation-
ship between things and is therefore independent of time. Thus,
ar ve jiyin, they went to him : their movement towards him was
permanent, not dependent on the moment, at any rate, as long as it
lasted. Similarly in the expression re vann ser, in country their, there
is obviously no temporal change going on. The one exception to this
is the preposition pitevi, for. Here pi is certainly the particle of the
future tense. In Uripiv the preposition is tevi, and the same word
seems to be seen in Eromangan tovnni. The rea.son for prefixing pi
seems to be that until a person has a thing actually in his hands it i.s
only '■ for ” him, the possession is still future, and logic bids one
indicate that future by the use of pi. This is perfectly clear when the
entire .sentence is future, but the feeling for a relative future can be
noticed even when the main idea of the .sentence is past, e.g. ar lev
nav'oni nono ya mu res pitevi, they took freely thing wh- it good (to be)
for him. i.c. they took a liberal .supply of good thing.s for him. He had
not received them yet ; they were still only ” for " him. The preposi-
tion is then a .sort of relative future even though the verb i.s past. On
the other hand the particle mn with res shows concomitance : the
things were already good when they took them. Other e.xamples are ;
Mare e tapnr vacin pitevi, the heaven was opened for him ; reyen
taronen mi mok rec pitevi, during the time he was still speaking to them
(these show a slight change in the meaning of pitevi. but onlv similar
GRAMMATICAL CATEGORIES IN’ THE NEW HEBRIDES
195
to that of the classical dative). In all these examples Uripiv lacks
the future prefix. So does Enghsh, and so do most languages ; yet
it cannot be denied that an event which has not yet occurred is still
future !
B. Time-words
In the treatment of time-words, we shall find it convenient to
retain the Enghsh nomenclature, viz., adjectives, adverbs, conjunc-
tions, and numerals, in order to show how the treatment of each of
these concepts in Achin sjmtax is such as to justify their being grouped
together under the category “ time- words “.
(i) Adjectives
In a number of Melanesian languages the adjective is always or
frequently verbahzed ; in nearly all it can be varied for person and
tense if used predicatively. Only in Achin, so far as I know, does it
vary systematically for number, person, and tense also, when used
attributively. Such a sentence as cinop rja mi res, a good man, is
fairly closely paralleled by the Mota o tanun we wia, where only the
ligative particle is missing. But as we say Everybody will hke
a good man’’, the Mota expression remains the same; the Achin
phrase becomes ci)iop ya pi res, the "adjective" becoming future
hke a verb. There are in Achin three uses of adjectives :
1. Attributively. Numbo toptap. sacred song [toptap = common
tapu). This use is not common, but extends to the numeral sa (Uripiv
san), one, when used indefinitely, e.g. cinop sa, a certain man.
2. In Indejinite Tense : the particle e precedes the adjective and
the ligative is not used : ivoiron vanu e lep. a big mountain. Change
for person can be made in this form, by the substitution of other
verbal pronouns for e : Ao worsnritn e pe bo res : I belie ve-you that you
are good (on e pe see under conjunctions). Po res e pe nise you shall
be good hke what ? i.e. How will you benefit ? This is chiefly a predi-
cative usage.
3. Other Tenses : In other tenses according to the verb of the
main clause. Par Ic.-a maciyen sami pn res : they shall see works your
(shall be) good — "shall be" because you have not yet done them.
Rnrpoy po roye rahtb pi mare : (in the) morning you shall hear my
voice (shall be) up above (but it is not there yet because I have not
yet cried out). Po kete lolomam pi ran, make our hearts (shall be) clean
(but they are not yet .so).
196
A. CAPELL —
In this last usage we see how the futurity of the main action is
transferred to the adjective also : the quality cannot be present until
the act which is to produce it is done. All this is perfectly logical, but
Achin has carried out the logic thoroughly ; the majority of languages
do not.
(ii) Adcerbs
The line of demarcation between adjective and adverb is not
perfectlv clear even in European languages, so that there should be no
ground for surprise that in Achin the two are treated similarly. Take,
for instance, a word that even in English is on the border line between
verb and adverb, the word ' over '. In Achin this is s>i. a word common
in Oceanic, and traceable back to Indone.sia. Xote how it i.s used in
the future tense : siveliy nawhoy sal: pa sa : when mv days shall be
over (exactly the same logic of tense as in French, quand mes jours
seront passes) : masaen e sa. the sickness is over ; iuasaea pu su, the
sickness will be over. Let this u.sage now be extended to words that
cannot also be verbs (though .some of them, of course, can be adjectives)
and the variation for tense becomes logical ;
1. Kop lei- pu res pi serser ini ya loJon e yar pitcvini : you shall
be good immediately he wh- his heart is bitter towards you, i.e
make it up immediately with him who is angry with you.
There are several points of intere.st here. («) Pu res is future because
the main verb, kop lek, is future ; (6) the action is to be taken serser.
immediately, but the whole thing is still future, and therefore the
action is viewed in the speaker's mind pi serser. future immediate.
The immediacy ’’ is yet to come ; (c) the prefix in pitcriai, which we
have already studied.
2. Fo royroyres pi lep e pe ya ko lek pu res ivor ne mare : you shall
rejoice greatly because you .shall be good (i.e. happy) in heaven.
Here {a) the adverb lep, greatly, is thrown into the future because
the main verb royroyres. rejoice (lit. feel good), is future ; it is a future-
joy with a future-greatness : .similarly (6) in the .second clause res,
good, is future because the being (lek) has not yet eventuated.
.3. hop van pecelei pu res /ipis ya mi ara war : go and .seek well
for the child wh-he is born.
They have not yet gone ; therefore the goodne.ss of the .search
i.s as future as the .search itself, and the fact is indicated bv the words
pu res. On the other hanrl. the child is alrcadv born ; hence the
tense particle mi.
GRAMMATICAL CATEGORIES IX THE XEW HEBRIDES
197
4. Po Jain pe viel pu res : lead me (and) I shall go properly. The
meaning here is ; I ivant to go the right way : lead me (future serving
as imperative), I shall go, and then it shall be good.
Just as there was an attributive use of the adjective, so there is
a use of the adverb without verbal particles : lolon e ar melker, he was
very angry. Again, there are border line uses with e : e ro e cac. he
was troubled (lit, he felt bad), as compared with the noun lolcac.
anger, ill-feeling. Similarly mar ln)j e lep e lep, they wept very greatlv.
An interesting case of a somew'hat similar nature occurs in the
largely Polynesian language of Futuna in the southern New Hebrides,
where the word weitrahu. immediately, is conjugated as a verb, to be
or do immediately.” Thus : aia iii iveiiraJia id visau, he spoke immedi-
ately, where the ni indicates the 3rd singular past tense. Thus the
idea of overlapping parts of speech is not absolutely limited to northern
JIalekula (indeed it could be illustrated from Rotuma and Fiji and
the Banks Islands and other parts of Oceania) and it would be
interesting from the viewpoint of psychology to collect detailed
examples from the syntax of other Island languages. In fact a com-
parative sjmta.x of these languages is greatly to be desired.
(iii) Numerals
In regard to numerals, too, there is more than one use in various
islands of Melanesia. We have already pointed out that the numeral
sa, one, can be used attributively in the indefinite sense. But normally
numerals are used, as in other languages, with the verbal particle
e : soponen e vac, part four. Now, once Achin admits a verbal particle
at all, it may bring in any verbal particle according to its sense of
logic, and so we get : pire cinop pi sa pi roije pu res Jco pi risenini
cinop e pe ok . . . : if any man shall be satisfied (lit. feel well about)
and shall teach men thus . . . Here the word pire, if, is followed by the
future tense, just as it would be in Latin in the same sentence (why
has French departed from this practice ?). and not be the hi-pothetical
particle ki, apparently for the same reason as in Latin, viz. that the
condition is possible of fulfilment. The whole idea is future ; not
only is the feeling of satisfaction future, but so is the man — he has
not yet come forward, and so the numeral is pi sa, “ future-one.”
Note again the future tense of the adverb pu res.
S. H. Ray (op. cit., p. 292) gives a similar example in Lvipiv :
pora ink ku loli uasi imnu pi san. if you can do amdhing (you have
not yet done it, therefore it is pi san. a future-one thing). Again :
198
A. CAPELL
iveya sa pi kanitn kop van roni maili jn sa, kop van roni ntaili pi yeru : or
if anyone makes you go with him one mile, you are to go with him two.
You have not yet gone either the one or the two ; hence, pi sa, pi yeru.
That this is simply a logical extension of the common Melanesian
uses of numerals as verbs, is shown not only by the use of e, but also
by the fact that the numeral can be used in the negative as well ;
wowon vanu e se sa te : there was not one mountain. There is another
particle ko, used with numerals, but I am not certain of its meaning ;
Nisei ya mu nac ko sa, another way ; mar ma re moral ya mi lep ko
sa, they came to a big place.
(iv) Conjunctions
If conjunctions can vary for tense, the explanation is most naturally
to be sought in the character of the words used as conjunctions.
More than one “part of speech”, even in English, can be used as
a conjunction. Thus we can say “ till he comes ”, or ■' till nightfall ”,
making one word do duty as conjunction and preposition. This is
true also of other languages, including Achin, with the result that in
Achin the peculiarities of time-words are extended to these derived
conjunctions. One therefore finds two classes of conjunctions :
1. Words that cannot serve any other purpose, or are object-
words and so invariable. Such is reya, in, while, which we have alreadv
studied. Another such is v:eya, or. AVe have also studied pire, if.
Another instructive example of this last-mentioned word is : pire pi
sa pi cova. pu cac pu ro : if there shall be anything further, it wdll be
absolutely wrong. The correlation of tenses is worth noticing.
2. Conjunctions that are really other " parts of .speech ”. A par-
ticularly important word of this .sort is pe, like. It is normally used in
a verbal form, e pe ok, like this, thus. As such it varies for tense :
pi pe ok. it will be thus. Combined with the ligativo ija. it serves all
the purposes of the English " that e.xcept its use in final clause.s.
Examples are : Mar ok e pe ya pnrofif e wire tua : the.se things are
as the prophets said of old. Po tikeir tipis wele ko miren e pe ya ini
ki pecelei tijns wele ok pi rerd pini : he will seek the young child and his
mother to kill him. Reya Rama mar roye e pe ya mar tay e lep c lep :
in Kama they heard that they were crying very greatlv ; e pe ya
Revel c tdysi natun wclek e se put mon te. e pe ya nafiin e rika vacin :
because Eachel was weeping for her children, becau.se thev were no
more ; ini pi pe ya mere Nasareti ko sa : He shall be as a person of
Nazareth (” He .shall be called a Nazarenc ") ; kap se kete te pi pe ok :
GRAMMATICAL CATEGORIES IN THE NEW HEBRIDES
199
don’t do it like that (you have not yet done it ; it is not yet “ like
that’ , hence pi ) ; po lot pitevi pi pe ya pi yoni, give to him as he
shall ask (he has not asked yet).
Sometimes, however, no conjunction at all is used in place of the
English “ that ”, as expressing either purpose or quotation ; Po kete
pe terter : Make I shall be strong, i.e. make me strong. Kami ko sa
e royivusoni pi kete turun po rav kele : Does one of you know (how to)
make his body taller again ?
Mention must also be made of methods of dealing with the
English relative “ when ”. In the future there is the word siveliy, an
example of which has already been given. In the past the word used
is van, come, and this can also be used in the future, according to
the time of the main verb. Thus : mi van e rorom lomtoni nono'k,
when he was thinking privately about these things, lit. it came,
he was thinking ... Pi van ko pi pep tmtun nortnan pi sa : when she
brings forth a male child, ht. it shall come (and) she shall bring forth
(future-)one child male. Note that the child is still a future-one,
like the bringing forth. Similarly, kop van kap pecelei pu res tipis
ya mi ara u'or, pi van kom sesewei . . . , go and seek well for the child
that is born, and when you (shall) find him ... So far they had neither
gone nor found him ; hence the future pi van, when. Incidentally
this sentence shows the root meaning of van as well as the secondary
meaning of the w'ord.
C. Parallels Elsewhere
Sufficient examples have now been given to justify the assertion
that the language of Achin requires the rearrangement of the parts
of speech into twm categories, object-words and time-w'ords. The
same could be said of Uripiv, though there the distinctions do not
seem to be so rigidly made. In the Fanting dialect of Ambrim there
are similar phenomena, as some of Ray's examples show. Thus on
p. 338 of Melanesian Island Lamjaagcs he says : ” The adjective
is very commonly used in a verbal form with the particles me. mi.
te. or be (i.e. the verbal particles) : vantin me hakabo, man bad ; rolo>t
mi yah, voice loud ; id mo lolo, clothes soft ; wobay be sul, dav (will
lie) third. Fanting has also the ligative in the form of yo : vantin
(fo mi mer : the dead man. One presumes that the same distinctions
of tense are made in the first three examples as is specified in the fourth-
though Ray does not state the fact, just as he does not bring out
the distinctions in Uripiv. We do not, however, find the extension
200
A. CAPELL
of tense here to adverbs,, conjunctions, and numerals, unless the numeral
be used predicatively, as in the example above quoted. It is note-
worthy that the southern half of IMalekula does not possess any of
these distinctive syntactic usages, though in both words and grammar
the languages show considerable resemblance to those of the north.
Apart from Ambrim, one can only point to a tendency in many
parts of ^Melanesia to use adjectives and numerals in a verbal form,
but there is no agreement in tense with that of the main verb, as here
in Achin and Fanting.
D. Interpretations
What conceptions of time and object he behind Achin syntax ?
This is an interesting question, but to answer it means “ thinking
black ", and that is a task which still remains for the white man to
accomplish.
Certainly here as elsewhere the unit of speech is not the word
but the sentence. We express in a sentence a thought that is already
complete in our minds. Emphasis is laid on various parts of the thought
in different languages. The Englishman is content to say “ that man ” ;
the Frenchman says <xt homme-la, and the Swede says den dar mannen,
mentally pointing him out (though they each do it differently even
here). Some languages emphasize the object of the action, some the
time of the action, some the manner of the action, and some other
features. The complications of the Amerindian languages show this
variation of emphasis advanced to a fine art. There is not only a tjqie
of mind that is unable to generalize (and the’’efore has separate words
for washing face, hands, feet, garments, etc.) and looks at things and
actions as distinct from others that to us are really of the same kind,
but there is aho a type that having baulked at these generalizations
proceeds to detail the exact circumstances of the particular act
mentioned. Thus in such a sentence as “ The man chased the boy
round the tree”, the form of the noun in Amerindian languages will
tell a lot more about the man and boy than in English, even to
stating whether the act took place in the speaker's sight or not.
In English we have simplified our statements to a very large
extent ; in the languages of the Far East, such as Tibetan and Chinese,
they are simplified still further. When we come to Melanesia we find
that although a con.siderable amount of simplification has taken
place (and one need only refer to the " Papuan languages and
many of the tongues of western Indonesia in contrast), still
GRAMMATICAL CATEGORIES IX THE XEW HEBRIDES
201
a good deal of definition is made that is not made in English.
Still, that definition in most places tends to be definition of concrete
things and acts. In the sentence given abov^e. the Melanesian would
tend to say " man that he chase him boy round tree " — i.e. the
emphasis is not on the time of the action, but on the persons concerned.
The act could be seen and the actors noted ; when it happened was
of less importance.
Now' in Achin and Uripiv we find the rev'erse tendency at work.
True, there is still the definition of the agent and the object- -the
verbal pronouns and demonstratives are still well in evidence —but
the time of the action becomes almost a ruling concept. That, of
course, is where the difficulty for the foreigner comes in. He has not
only to learn native words and idioms, but has to reorientate his
methods of thought along the lines of native methods. He has to
throw the emphasis where the native would throw it. In the language
we have been studying this emphasis is on the time element.
What is the effect of this emphasis on time, and w'hat guides its
application ? In the first place, the only things independent of time
relations are those that persist unchanged throughout all experience.
The man who acts is the same before as after ; so his name “ man ”
is independent of time. The boy who is chased is still a boy afterwards,
so his name “ boy ” is unchanged. If, now', w'e extend the sentence by
adding the words “ till he caught him ", we add elements that can be
affected by the time emphasis. “ Till ’’ is not the same in the given
sentence as in the sentence “ The man will chase the boy round the
tree till he catches him ” — to us, perhaps, it is, for all practical purposes,
the same, but not to the native. In one case he says, “What happened ? ’ '
— why, he chased him, “ and it came he caught him ” ; in the other,
“ What will happen ? ” — why, he will chase him, “ and it will come
he will catch him.”
Take another t}’pe of sentence. The Matthsean text “ if thine
eye be evil, thy whole body w'ill be evil, and thou shalt be full of dark-
ness ” has been rendered into Achin as follow's :
Pire metom pu cac If thine eye shall be bad
inik pu su pu cac thou (shalt be) all (shall be) bad
pu ror pi lep. it (ivill be) dark (will be) great.
From the Enghsh point of view this means nothing, but in the native
mind it must point to a definite analysis of the whole idea, made before
the idea has been expressed. The idea is the connection between
202
GRAMMATICAL CATEGORIES IN THE NEW HEBRIDES
mental and moral darkness. Granted, of course, that the idea itself
is foreign to the native mind, how is it to be put into his mind bj the
missionary whose business it is to make him understand it ? In the
first place, the thing is certain, though it has not yet come to pass.
But it can — it is merely future, not entirely hypothetical. The idea
will require the use of the future particle (pi) not the hypothetical (L’i).
Now, what parts of the idea can be affected by this contingent futurity ?
Obviously not your eye — you have that already. But the badness is
future, and so is the darkness that will result “ if thine eye be evil
Then the e\fil is still future, as well. That is as far as the Classical
Languages would go in the analysis, and farther than English and
many other modern languages would press it. But the Achin native
seems to say, “ Well, if your eye is evil, will it affect the whole or part
of your life ? ” If the w'hole, then that is part of the futurity, and we
get mil- pu su, you will be entire(ly), i.e. the whole of you will be evil.
Then there is the darkness, jni ror, it will be dark ; yes, but it is not
yet dark, ex hypothesi. Then the greatness of the darkness is also
a future affair, and we get it expressed in the future particle pi lep,
it will be great. If it remains true that the unit of speech is the sentence,
it seems true also that the sentence itself is not identical in every
language (the Latin “ periodic construction” and the English, say, of
Macaulay is a case in point), and it would seem that to the Achin
native each section of the idea becomes a separate sentence : pire
metom pu cac — inik pu su — pu ror — pi lep. The words and phrases
hang together (and this may be the psychological ground for the
changeable vowel of the verbal particles), but all that connects the whole
idea is the futurity of the matter. In fact, we have, not really an
elaborate logic carrying the idea of futurity to its utmost possibility,
so much as a very elementary thinking in disjointed ideas, each
expressed separately, and only bound together bv the future particle.
This is only a tentative suggestion of what may be the native
outlook, and may be open to correction, but it is at least highly
interesting as an unusual development of the time concept in native
speech.
REVIEWS OF BOOKS
Linguistica
By R. L. Turner
L'Indo-Aryen DU Veda aux Temps Modernes. By Jules Bloch.
X pp. 335. Paris, Adrien-Maisonneuve, 1934. Frs. 80.
The ‘ discovery ' of Sanskrit by Europeans, and in particular by
Sir Wniiam Jones, did more than any other single factor to stimulate
and nourish scientific interest in that comparative and historical
study of the Indo-European languages, which alone has formed the
firm basis for the training of comparative linguists whether in that
or other fields. All Indo-Europeanists have of necessity studied
Sanskrit, which till the discovery of Hittite, presented the oldest
documents of any Indo-European language : indeed in the first
enthusiasm the earlier linguists painted a picture of the ancestral
Indo-European language that was probably far too Sanskritic in its
colouring. It is therefore curious that a language so important to
the linguist should have had till quite recently so little attention
paid to its subsequent developments. Yet these present views of the
greatest interest to the comparative and historical linguist. From
Sanskrit is descended the vast family of languages spoken by over
270 milhon people from the borders of Baluchistan to the Bay of
Bengal, from the Hindu Kush to the 15th degree of latitude far south
in the Deccan, and outside India by the Singhalese in the island of
Ceylon and throughout Western Asia and Europe by the wandering
and settled tribes of gypsies.
Several of these modern Sanskritic languages have long literary
histories or are documented by early inscriptions : Bengali, Hindi,
Gujarati, Marathi, Singhale.se. A process of great importance in the
history of language, namely the extension of a dominant language
over other linguistic areas, can be observed taking place even now.
The history of Sanskrit and its descendants can be traced from perhaps
the middle of the second millennium b.c. to the present day, a period
longer by nearly a 1.000 years than any similarly documented period
in the history of any other language-group. Meillet in his incomparable
Apergu de la langue grecque and later in his Esquisse d'une
histoire de la langue latine showed what interest for the general
204
REVIEWS OF BOOKS
reader and scientific profit for the student may be had when a master
of linguistic science displays the main lines of the development of
a single language-group and the chief influences in its history.
Nurtured in the linguistic school of ileillet. an Indianist trained
by Sylvain Levi, Professor Jules Bloch, himself the originator of the
scientific comparative study of the modern Indo-Aryan languages
in his famous Formation de la kingue tnnrathe} has now in this
long desired volume described with an astonishing conciseness, and
vet with a great wealth of detail, the main developments of that
Indo-European dialect, of which the first document is the Rgveda
and the present spoken forms are the modern Indo-Aryan languages.
The whole book is informed with so personal an insight into
the problems, so critical a linguistic sense, so just an appreciation of
the different factors of development and throughout so scientific
a spirit, that no linguist, whatever his particular field, can fail to
profit by its reading, no Indianist. whether comparativist or not,
can afford to be without it.
Many problems, both major and minor, as Bloch throughout his
work points out, still await solution. There is for the earlier history
of the modern languages a wealth of material still untouched by the
linguist. Here are a few points. The apparent absence of a form
corresponding to Skt. tfna- in Iranian is adduced to support the
argument that Finno-Lgrian (with its Ostiak toran) borrowed some
at least of its Aryan vocabulary from an Indo-Aryan rather than an
Iranian source. Yet *trna- is attested for Iranian by Saka ttarra
‘ grass, straw ' (Konow, Saka Studies p. 187).
It is often tempting to see survivals or archaisms in forms which
can perhaps be better explained. Pa. gahita- (p. 14) in my opinion
is more likely an analogical replacement of *ga}ilta~ < Skt. gr(b)hltd-
(given the exceeding frequency of the ending -ita-) than a survival
of a Prim. lA. *grbhita-. The assumption, with Helmer Smith, that
mnha- (p. 15) is a survival of an IE. form beginning with u-, not yu-,
is unnecessary since (BSOS. V. p. 46) initial y- in pronominal forms
is liable to disappear (Asokan a- etc. < gn-) and the Yajnavalkya
Siksa prescribes for the v- of rah and vdm- a light pronunciation.
^ The first and immediate success of this work has had one unfortunate result.
Too many students have been led to think that despite inadequate material
or preparation they could do the same for other modern Indian languages. May
such in the future turn rather to the solution of the many problems which
Professor Bloi h i^ugge^ts for study in the present volume.
LINGUISTICA
205
It has been generally assumed that Pa. vlsati (p. 37) represents IE.
*ivl- (Lat. vlgintl, etc.). But is it necessary to go bevond Skt. vimsati- I
For iinh already in Pa. > Vi (slim- < simhd-) as later runh in Pkt.
> dll ; tins > Pkt. Is (pis- 'grind ' < pit'ns-). although remaining in
Pali. But sound-change appears earlier in numerals than in other
words ; witness As. traidasa < Irdyodasa and cdvudasam with earlv
loss of -t-. Thus the change, ims > Is, is in accordance with the
system and appears in the first favourable circumstances.
On the contrary in another in.stance the survival of an ablaut
form not found in Skt. is likely. Bloch (p. 56) following other writers
assumes that in three words Skt. palliati, Pa. patliaum- and pnti-
an original dental occlusive has been cereliralized by a preceding r.
although not in contact with it, and derives from pratli-. prathnmd-.
prdti. The fourth example given is Pa. sathiJa- beside Skt. sitliild-
and Pkt. sidhila- as belonging to Skt. sratli-. Herein surely lies the
solution of the problem ; sithild- etc., as the variation of vowel shows,
is derived not from *smthila-, but from *stthild- (Acp. Diet. s.v. dhilo)
with the vocalization proper to this formation. Similarly may we not
assume *prtMti heside prdthate (cf. prfltu-, and the tj'pe tirdti : tdrati) as
origin of pathati (despite the gram, accent pdthaii). This is indirectly
supported by Pkt. appdliai ‘ teaches ‘ < *dprdtka>jati beside padhai
< *prthdti. Similarly patliaina- < *prtlia>m- beside pratliamd- (with
the same vocalization as OHG. /ordar beside Gk. rrpoTepos); and pati
< *prti (as in Geim. fort ; cf. the coexistence of the two ablaut-forms
in one language-group, Gk. npos and irpcs).
Skt. final vowels survive as whispered vowels more e.xtensively
than is indicated on p. 42. In Gujarati, as N. B. Divatia (Gujarati
Language p. 224) and T. N. Dave (BSOS. VI, p. 677) have shown,
MidlA. -l survives as a whispered vowel or a palatalization of the
preceding consonant. But this is not shown in writing. Against
this general reduction or disappearance of final vowels it is difficult
to admit (except for monosyllables) Bloch's contention that in
exceptional cases Skt. final long vowels remain ; and I prefer to derive
Mar. dinlil. not from Pkt. amhc (p. 43), but from the inst. amhehim :
the nasal is retained in Pj. and Si. asi. where it is not ma.sked bv a
preceding m. The replacement of the nominative by the instrumental
in the personal pronouns is common (t\'pe Hi. nml) ; and it is at least
noteworthy that in the 1st plur. OGuj. -aum < -dnmli (e.g. jdnaum
‘ we know ’) has been replaced by what was originally a passive
construction demanding a pronoun in the instrumental, arm jdrnge
203
REVIEWS OF BOOKS
(= Pkt. amhehim jdmai), which is left unexplained (p. 248). A similar
invasion of the passive probably accounts for the 2nd and perhaps
3rd sg. of the Nep. imperative-subjunctive (p. 246) gares, gare (or
garyes, garye).
Amidst all the detail, especially in the chapters on Phonology,
there are naturally some points on which all will not find themselves
in full agreement with Professor Bloch. The following, however,
which I view somewhat differently from him, do not for the most
part affect the main lines of his demonstration or the general accuracy
of the picture he set out to draw.
P. 18. It would be better to speak of the passage of t rather than d
to r in Khowar. For Skt. -d- disappears and only -t- remains as -r- :
chan ‘ leaf ’ < chadana- (Morgenstierne, Report ^ p. 72), uy ‘ water '
< udaM- ; similarly -th- > r, hut -dh- loses its occlusion chui
( = chui ?) ‘ hunger ’ < ksudha, prat ‘ he beat ’ cf. pradhat, gom
‘ wheat ’ < godhUrna-, etc. Kalasha should be included as a language
in which > -I- (hence derivation of tara < Skt. tatah suggested
on p. 198 is suspect). Shina, included here and on p. 63, must be
definitely excluded (see BSOS. IV, p. 533 ff.). And Kalasha, like
Khowar, loses the occlusion of -d- as opposed to -t- : uk ‘ water ’,
uisau ‘ boils ’ < *uiml < udisali ‘ rises ’.
P. 19. Kalasha is to be cla.ssed rather as Dardic than as Kafiri
(Morgenstierne, Report^ p. 51): the strong Kafiri element in its
vocabulary is almost certainly loan material from the adjoining
Kati {ih. p. 52).
P. 32. The MI. development of r to {r)u only in the presence of
labials is not quite comparable with the distinction between Skt. ir
and ur, for though the latter represents IE. °r (or better «r) in the
presence of labials, it also appears where there is no labial and represents
rather IE. °r : firm- but dturta-, tirdti but supratur-. *pratnrati {Nep.
Diet. s.v. panrann), tlrlhd- as well as *turtha- (Nep. Diet. s.v. turnu).
P. 33. Is it correct to say -o issu de *az decompose parfois en
-ay " ? Bather *az before initial vowel > ay. une.xploded intervocalic
> !/> just as later in Gilgiti Shina (from Skt. -s-) > -y- : haydiki
‘ to laugh ’ < hdsayati. sdyure ‘ sisters ' < .smsdrah beside Kohistaui
Shina hazonu, sazdre. It is this -ay which becomes -a before vowels,
as -dy (< *di) became -d.
P. 34. Stability of sound-.system as between Sanskrit and ModlA.
applies only to the languages of India proper, which were most
influenced by and themselves influenced Sanskrit as a literary language.
LINGUISTICA
207
On the borders, particularly in the North-west, the system was far
less stable, and a number of new phonemes appeared : e.g. the voiced
correspondents of s s s, namely z z z, the cerebrals c ch j, the spirants
X y h f, while the groups containing r show a bewildering variety of
soimds unknown to Sanskrit.
P. 35. It is not clear how Skt. srnoti proves indirectly the existence
of a treatment r -1- vowel, since it correctly represents an IE.
*]d-ne-u-. On the other hand Asoka sruneyu etc. probably owes its
rii to the past participle snitd- < IE. *kluto-. Pa. pucckati, vicchika-
etc. are ambiguous, for they may well represent earlier *prucchati,
*mcchika- : there is therefore no necessary contrast between Dutr.
prudhi and Pa. puthu (prthak) except that in the N\V. form pr survives,
while in the East even before the time of Asoka pr- > p- ; ef. the case
of Pj. rrMter ‘ stepmother ’ < *maltiara < *mdtritard < *mdtrtard
discussed below, p. 220.
P. 40. Anusvara before a consonant indicated something more
than a mere nasalization of the vowel, for when that something was
lost, the nasalized vowel was lengthened {vamsd- > has), just as when
the consonantal element of n in the group nt was lost {ddnta- > ddt).
P. 47. In face of the nasalization of vowels in the presence of a
nasal consonant, we may note that in OGuj. the ending of 3rd sg.
pres, -ai ( < -ati) becomes -aim if the verb is preceded by the negative na.
P. 49. Eead dvdra- rather than duvdra- as the source of Shina
ddrl ‘ doors ’.
P. 53. The Kafiri dental semi-occlusives cannot be compared
with those of Kashmiri, Singhalese etc., for the latter represent Skt.
c. j, while Kaf. c corresponds to Skt. s, and Kaf. has only palatals
c. / corresponding to Skt. c, j. MI. jhd- as source of Ksh. zdl is a slip
iovjdl- jvdlaynti) as correctly given later. Ksh. chuh ' is ’ presents
no phonological difficulty, since it is < Skt. dkseti, with regular ch
< ks [BSOS. V, p. 137 ff.).
P. 54. In Sindlii jn regularly appears as n : ohdnu < *upajndna-.
vlMnii < rijndna-. In janyo < ynjndimntd- the dissimilation of n
to n (due to the other palatals) was later than the simplification of
double consonants, since MI. nn > Si. n. In < djnd and-m/iT <
rdjhl the phenomenon is Middle and Common Inlian, retention of
the long vowel d through analogy (tj'pe d-sarati etc. and masculine
rdjd) led to the early .shortening of the following double consonant.
Guj. and Mar. equally have dn and rani, not *dn, *rdnl.
P. 55. Is it not possible that Asoka Girnar das(s)ana- with n
208
REVIEWS OF BOOKS
may as an administrative term be a loan from the Eastern dialect,
for whose earlier change of -n- > we have some evidence.
P. 56. The confusions of dental and cerebral development of the
group r and dental are due not to confu.sion or hesitations in phonetic
change, but are facts of vocabulary, due to the extension of particular
words from one dialect to another. “ Le fait dominant " is not so
much “ I’extension nouvelle des cerebrales ", as the extension of
particular words containing cerebrals.
P. 57. In Shina tr > c, not I : hence got ‘ house ' < goMhd- (not
gotrd-) as in Tir. gusta (Jlorgen.stierne AO. xii. p. 180), and other
Dardic languages ; and patu ‘ leaf (unless a loan from a dialect of
the Garwi tj-pe) is < patta- rather than pdtfra-, whether or not there
is any ultimate connection between these two {Sep. Diet. s.v. pdt^).
Skt. pdtati remains in all the N\V., \’iz. Sindhi, Lahnda, Panjabi
and West Paharl dialects, besides Kashmiri.
P. 62. It should perhaps be noted that in Mod. Singhalese h
represents earlier Sgh. s < Skt. c, while earlier Sgh. h < Skt. s has
disappeared : thus dm ‘ wild boar ’ < sukard-, but hord ‘ thief ’ <
cord-. The difference in the dates of these developments is clearly
seen in the inscriptions.
P. 63. Kati fta ‘ given ' and corresponding Dardic forms on
p. 270 from prdtta- (Morgenstierne, Report ^ p. 53) rather than prapta- ;
cf. also Kalasha pra ' I gave ’ < pradarn beside present deoi.
P. 65. In my mind there is no doubt that Pa. kdhatni is derived
not from *kar.sydmi but from *kd(s)sdmi formed after dd(s)sdmi and
ddhdmi < ddsydmi, just as Pkt. kd(d)urh is not derived from
kdrtum, but formed after datum etc. (cf. BSOS. VI, p. 531 ff., where the
influence of the verbs in -d on MI. conjugation is discus.sed).
P. 67. The Guj. form is ame. not home, and does not therefore
come into question here.
P. 70. It is incorrect to say that Asiatic Gypsy has only one
sibilant to represent the three of Skt. Unlike European Gypsy it
confuses s and s as .s. but maintains s as s : only Armenian Gypsv
confuses all three. Cf. my' Pn.’iiliou oj Romuni in Indo- Aryan, p. 19 ff.
P. 71. *drlgg}i(t- (as on p. 85) rather than *drlgha- as the inter-
mediary' between dlrghd- and Shina Agu.
P. 81. In regard to Asoka Girnar thaira- < sthdvirn- beside nsti
< dsti. there would be nothing surprising in a difference of treatment
between initial st(h)- and medial -st(h)-. since this is seen in Gypsy
(Position of Romani, p. 22). On the other hand itthi with assimilation
LINGUISTICA
209
both of st and of r is clearly an Eastern loan, since Girnar maintained
tr ; at the same time thaira-, later used as a technical Buddhist term
and already a term of address as shown by the unexpectedly early
loss of -V- and crasis of resultant ai, may equally well be an Eastern
loan. Although Girnar apparently confused st and stJi at least in the
intervocalic position, this was not universal, for Kalasha distinguishes
initial st- > ist- from sth- > th- : istonim < stdndmi, istri < stri,
but them < sthapdydmi, thur < sthnnd, thul < sthuld-.
P. 83. The forms of Nep. pdu- and Guj. -pdm- do not confirm MI.
pdpun-, since they derive not from this but from prdpayati. On tbe
other hand the stem of the Sgh. verb is not pcim- but pdmin-, and
this like Dutr. pamuni rests on prapnoti > *pamn- > pamun-, while
Gir. prdpundti represents insertion of vowel before -pn- > mn, unless
as is possible -m- in both Dutr. and Sgh. represents -v- < -P-.
P. 84. Whereas dvdra- shows two dialectical developments ddra-
and bdra-, dvlpd- has one only as ddpa-, for v was dissimilated by the
following labial, just as we find early loss of -v- in Pkt. in words
beginning with a labial piai < pibati, paisa'i, uvaisa'i < prdvisati,
upavisati.
P. 86. Ksh. wanda, Shina ydmi ‘ winter ’ cannot certainly be
derived from kemantd- since both languages preserve Skt. -m-. Are
they < *hemna- ? Cf. Skt. hemein in winter = Ashkun ze, Dameli
zln (in both of which the -nt- of hemantd- would remain).
P. 91. One important soimce of the gemination of consonants is
overlooked. A language possessed only of double intervocalic con-
sonants when it borrows a foreign word containing the unknown
single intervocalic consonant naturally reproduces it as the known
double. It appears therefore by no means improbable that ML. know-
ing onlv double consonants and borrowing from the learned language
a word like eha-, could reproduce it only as ehha-, in the same way
as Mod. Panjabi doubles the single intervocalic occlusives of Sanskrit
or other foreign loanwords : Pj. dttd with long d (opposed to Si. ato)
is loan from Hi. dtd. On p. 93 the single -s- of Si. asl is not reallv
opposed to the double of Pj. assl, since Si. asJ rests on older assl
with the same doubling ; ^H. asli < Skt. aslti- would have appeared
in Si. as *ahl.
P. 123. It is difficult to see how stems in -in- influenced stems
in -i- so as to produce acc. pi. in -~in. Is not this rather formed on the
analoffv of the masc. stems in -n- : -nm : -dn = -un : -In, while -d)i
for *-ons is due to the analogy -ah : -dh = -am : -dn I
V<II. Vlir PART 1
14
210
REVIEWS OF BOOKS
P. 143. I have maintained elsewhere (JRAS. 1927, p. 232 If.)
that Pkt. gen. sg. -aha is < -asya, of which the form -asa is attested
by Gj’psy, and -am ( = -aza) by the Khar. Doc.
P. 144. The fact that Apabhrarhsa distinguishes a locative case
in the singular but not in the plural, is by no means necessarily a
sign of its artificiality. Por distinction of case seems to be a need
felt more for the singular than for the plural. As is well known, IE.
distinguished dative and ablative in the singular, but not in the
plural. The terminations of the dat.. abh, inst. and loc. plural with
their greater variety of form indicate later creation in IE. than the
corresponding terminations of the singular ; in the dual we have no
evidence for a distinction of more than three cases. And in fact in the
modern lA. languages separate forms for both loc. and inst. have
lingered longer in the singular than in the plural.
P. 151. It seems unnecessary to follow A. Pillai and L. V. R.
Iyer {Ed. Rev. Ahtdras, Oct. 1928, p. 6) against Caldwell and Gundert
and to assume that the final -am of Tamil and Telugu nouns is borrowed
from Skt. neuter nom. aec. -am. For their view neglects the fact that
this -am appears also in the plural in Kanarese before the suffix -gal.
P. 152. B. refers to the general use of a postposition to replace
the direct case (old accusative) with animate nouns. This differentia-
tion of animate and inanimate declension showed itself first in the
use of the oblique case ( usually the old genitive) to replace the accusative
of animate nouns ; cf. the similar process in Slavonic in which also
the old nom. and acc. had coalesced and the old genitive is used as
the direct object case of animate nouns. This use of the oblique is
still found in Gypsy, where -es (< -asya) sometimes by itself forms
the direct object case of animate nouns, the old nom. acc. or ‘ direct ’
case being used in a similar function for inanimate nouns.
P. 153. Hi. tdrd m. is probably an inherited form, not borrowed
from Skt. tdrd, cf. Skt. tdrala- n. in MBh. : for there is no reason why
Nep., Kum., Bhad., Si., Guj., all with Idro m., should have conceived
of a Skt. loanword tdrd as masculine, since in these languages the
nom. sg. masc. ends in -o, not -d.
P. 173. Kashmiri siitin ‘ with ‘ < *sakte)in rather than *sahitena,
which does not account for Ksh. t.
P. 186. In Lahnda the comparative suffix -erd < -a-tara- still
remains (O'Brien, Gram. 3Iullam. p. 16).
P. 199. To refuse the development -s- > -h- in certain conditions
for nearly all lA. languages is to make unnecessary difficulty. That
LIXGUISTICA
211
-s- in a pronoun (or in a termination, JRAS. 1927, p. 230 tf.) should
become -h- earlier than in other words holds nothing surprising, when
the phonetic history of unstressed pronominal forms in whatever
language is considered. There appears to be no objection to deriving
Apabhrarnsa aJio, eho from Skt. asau, esd (Pkt. eso). So with another
pronoun Si. c^d ‘ what ? ' which B. wrongly connects with Skt. hm. not
accounting for Si. ch. Guj. so beside earlier kaso < kisiu, and Beanies
derivation from *ksd < *kasa suggest its origin from Skt. kldrsa-
> *kiisa- > *ksa- with loss of the interior vowels before -s- > -h-.
P. 210. The grounds for assuming an lA. *diddti beside dddati
are slender. The i of Pkt. dinna- and Pj. dittd is to be ascribed rather
to the influence of the past part, -dita- seen in vyadita-, Hi. diyd etc.
(Tedesco JAOS. 43, p. 358 If.) and indirectly attested in prdtta- etc.
P. 237. The evidence for the existence of the Skt. augment in
past tenses of Khowar and Kalasha is passed ov'er too hurriedly.
■ Irregular ’ forms are more decisive, like Khow. blr ‘ he goes ’ : boyai
‘ he went ’ (cf. Skt. vyHi : rydgdt), or Kal. plm ‘ I drink ’ jitn ‘ they
drink ’ : awis ‘ I drank ’ aun ‘ they drank ’ (as given by Leitner :
LSI. has also the newer analogical forms apis etc. : Skt. pibdmi :
apibani), and Kal. kdreu ‘ he does ’ : dro ‘ he did ’ (cf. Skt. karoti :
dkarot). In the face of the MI. change of consonants made intervocalic
by the addition of the augment, it is difficult to see anything but the
survival of the augment in these forms (BSOS. IV, p. 538 ff.).
P. 267. The obi. pres. part, in -td in Guj. and Mar. refers to the
subject of the main sentence : e.g. to mild mild khdll padld ‘ he fell down
while walking ’. But in Old Guj. it never referred to the subject of the
main sentence, but was of the type leha naim jlvatdm sukha ‘ to them
while alive there is happiness ' or seratdm vimdsivauih ‘ while people
are using, it should be considered ’. The origin then appears to be
a genitive plural of the semi-absolute t^-pe found in Sanskrit.
P. 283. The statement that the Sanskrit infinitive has disappeared
except perhaps in Marathi, is too sweeping. It survives still in Hindi
and Gujarati with the verb “ to be able " : mal kar saktd hu or hu
karl saku chu reproduces the Skt. t}'pe saknomi kartum. MI. karium
> karl, kar equally with MI. karia.
P. 285. Asoka karum ‘ while doing ’ can equally well be
a nom. sg. pres, participle replacing kurcan (-m < -n) and formed
on the stem of karo- kuni-. It is in just such a common word as the
verb “to do " that the old form might resist, or exist side by side
with (as here) the new analogical formation in -anta-.
212
REVIEWS OF BOOKS
P. 308. To the four languages in which alone md is said to survive
should be added Kafiri (Daineli tm), Dardic (Kal. m-o, Kho. ma),
Gujarati ma, ma.
The great contribution which Professor Bloch makes is that
amidst all this detail much of which is uncertain and must remain so
till far more workers have entered the field, he has produced a clear
picture of the main line of dev'elopment undergone bv the Indo-
Aryan languages and has displayed wherever possible the system of
that development. This is equally true of the section on morphology,
in which Professor Bloch has made even greater personal contributions.
A Eomax Alphabet for India. By SltsIti Kumar Chatterji.
Calcutta University Phonetic Studies, No. I. X 9|^, pp. 58.
Calcutta University Press, 1935.
A single alphabet for the whole of India, in which all its languages
should generally be written and printed, is greatly to be desired.
Were it not for the use of the Perso- Arabic script by large communities
of Moslems and others, the Devanagarl script might have served this
purpose, and have replaced the other descendants of Brahml. But
there is little hope of the general adoption of a Nagarl script by
Moslems accustomed to the Perso-Arabic. Further, it would seem
desirable that the main Indian languages, such as Hindustani or
Bengali, which will be the vehicles of Indian culture to the rest of
the world in the future, should not place any necessary obstacles in
the way of foreigners' acquisition of them. No doubt even Russian
and Russian literature would have been more widely known in Western
Europe but for its unfamiliar alphabet. The Roman alphabet has long
been used for transliterating Indian scripts, but the larger number
of symbols required for these much more phonetic scripts has neces-
sitated the use of a number of diacritical marks usually placed above
or below the letter concerned. These marks are not generally
available in Indian pre.s.ses ; they are liable to damage in tj’pe, and to
be omitted by careless writers or compositors.
Professor Chatterji in this most interesting, and by far the most
practical, of systems based on Roman avoids the first two difficulties
firstly by using no new letters or other symbols not avnxilable in all
founts, and secondly by putting such diacritical marks as are em-
ployed on a line with, and not above or below, the letter concerned.
The new letters reciuired he obtains partly by diacritical marks such
as ; after a vowel to denote length, ' after a consonant to denote
LINGUISTICA
213
cerebral pronunciation ; partly by the use of some inverted letters.
This has the very great merit of reducing the number of diacritical
marks required and of entirely avoiding the necessity for new tj-pe.
But it would be helpful if the author indicated what form these in-
verted letters should take in handwriting. For a few languages
needing additional symbols, italic letters are suggested. In my opinion
it is a mistake to mix t^-pes : it is ugly ; it would be a constant source
of confusion in composition ; and it would be difficult to make the
distinction in handwriting. Perhaps Professor Chatter ji could avoid
this by some other device. The pamphlet deserves to be studied,
and its main principles adopted, by men of all parties and languages
and religions in India.
Vedic Variants. Vol. Ill : Noun and Pronoun Inflection. By
M. Bloomfield, F. Edgerton, and M. B. Emeneau. 6| x 9|,
pp. 513. Linguistic Society of America, Philadelphia, 1934.
The value and special bearing on Vedic philology and the linguistic
history of Indo-Aryan of the great work undertaken by Professor
Edgerton in ‘ Vedic Variants ’ were enlarged upon in reviewing
the first two volumes. The author has now given us the third
volume of this study of the variant readings in the repeated mantras
of the Veda. In this Edgerton's own contribution is even greater
than in the previous volumes ; for here Bloomfield's work did not
go beyond a preliminary and incomplete collection of materials ;
and although Dr. il. B. Emeneau has given most valuable assistance,
the responsibility for the final form of the work is entirely Edgerton's.
The material, the variants in noun and pronoun inflection, here
analysed with such great accuracy and insight falls into two parts.
The first dealing with formal variants is reallv an addition to the
descriptive grammar of Vedic, which supplements and occasionallv
corrects the third volume of the Allindische Grammatik. The second
part contains the syntactic and stylistic variants. These, also, reallv
fall into two separate categories, of which the first, the S}Titactic
variants, has the greater bearing on the history of linguistic develop-
ment, The quite surprising interchange of cases which is disclosed
by the variants throws much light on the history of a language which
in its spoken forms was progressively decreasing the number and the
clear distinction of its cases, and which has resulted in some modern
languages practically devoid of nominal inflection at all.
In Dr. Emeneau, Professor Edgerton has an assistant who will
help him to complete the publication of the great work, of which the
214
KEVIEWS OF BOOKS
first tliree volumes, so extensive in tlieir material, so well arranged in
their presentation and so meticulous in their accuracy, have already
astonished us with the speed of their production.
ZUE IXDOIEAXTSCHEX UNO GEIECHISCHEX XoMIXALBILDUXG. Bv
H.JALiiAR Frisk. Goteborgs Kungl. Vetenskaps- och Vitterhets-
samhalles Handlingar, Ser. A, Band 4, Xo. 4. 6f X 9£, pp. 75.
Blander, Goteborg, 1934.
Dr. Frisk has made a thorough study of all Sanskrit words containing,
or appearing to contain, the suffix -ra-. Starting from the comparatively
small number of forms shown by exact equivalents in two or more
other IE. languages to be actually received from Indo-European, he
shows the continued growth of formations with this suffix and its various
derivatives. The whole is informed with a severe linguistic and philo-
logical method in keeping with the quotation from Meillet's La aiethode
comparative en linquhtiqxe historique with which F. prefaces his work.
A few points. The identification (p. 18) of vapra- ‘ mound ’ with Av.
vafra- ‘ snow ’ proposed by Morgenstierne AO. i, 256, is held to be
somewhat doubtful on accoimt of meaning. But it is supported by
the similar meanings of Skt. kiri- in dkhukiri- ‘ mole-heap ’ and among
modern Danlic languages as recorded by Morgenstierne : Kalasha
'klrik ‘ snow Palula klr {NTS. v, 339 where M. compares Skt. kirdti),
Dameli klr (Report^, 58). tamrd-, though d;7. Aey. in RV., possibly
survives in Giqisy Greek tarn ‘ blindly ’ with an extension in Welsh etc.
tamlo ‘ dark, obscure '.
This study is followed by .some equally careful and well-informed
smaller articles ; — 1) on RV. Vnibhuva'id-, which F. shows is a haplology
for *vaibhumsava-. derivative of V ibhitva-'in - ; 2) on the formation of
Bahurahi compounds in Old Iranian of which he displays the tendency
to form thematic stems ; 3) the existence of an -uk- infix in Greek and
Indo-Arvan.
Le.S FoRMATION.S XO.MIXALES ET VERBALES ex P DU S.\NSKRIT. By
B.xtakrishna Ghosh. GI x 10. pp. 114. Adrien-Maisonneuve,
Paris, 1933. 25 f.
On the suggestion of Professor Wiist, Dr. Ghosh has studied with
profit the words in Sanskrit which suggest the existence of a nominal
suffix -pa- and at the same time the so-called causative suffix -paya-.
He has made out a good case for the existence of such a nominal suffix in
Sanskrit, although there are few traces in other IE. languages. I would
refer especially to his analysis oMkitpa-, stupa-, puspa-, sd.spa-, tdlpa-. The
LINGUISTICA
215
-jp- oi-paya- Dr. Ghosh considers, and produces good reason for considering,
an enlargement of the root, without at first any specific causative
meaning : the later attachment of definite causative meaning to it is
paralleled by the history of the suffix -aya-. Some points perhaps
require further discussion or elucidation. In making the, to me at least,
highly doubtful suggestion that the n of slond- and shvnna- in face of
srond- and srdvana- is due to a cerebral pronunciation of 1. Gh. has
neglected the suggestion of the late Professor Charpentier that the n of
lavand- was so changed before r> I, and that lavann- therefore pre-
supposes a *mvana-. To say (p. 15) that r alternates with «/ in ,y- :
*sal- (in ucclial-) is surelv wrong ; rather ar alternates with al {mrati :
*salat{, like cdrati : calati). The grammarians' dhdtu still provides
pitfalls for the unwary. In showing that puspa- is to be analysed as
pus-pa- Gh. has omitted pusya- of which the meaning ‘ flower ' (cf.
also AV. pusya f. ‘ a species of plant ’) is attested in the modern lan-
guages ; nor does he mention the theory of Helmer Smith that pitspa-
represents a NTV. form of *pusma-, by which would be explained the
pusa of the MS. Dut. du Rheims. The omission of pusya- is the more
surprising as it provides an excellent parallel to the other pair discussed,
saspa- and sasyd : for according to Gh. saspa- is < *sas-pa-. Given
the change of initial s- to s-, for which, however, Gh. offers no satisfactory
explanation, there is no need to assume the work of analogy for the change
of s to s. That may be purely phonetic. In Sanskrit, as is well known,
Ar. s — s>s — s (h'dsura-), and s — s>s — s {h'tsyati). Similarly
s — s> s — s (saM-). But just as the assimilation s — s> s — s
W’ould lead to the unusual initial s- so that it was replaced by s-, so also
in *saspa- the assimilation of s — s to s — s would lead to the unusual
combination -sp-, and it was replaced by -sp-. It is wrong to say (p. 57)
that, in face of the older kasyapa-, the etymology of Yaska [kacclmm pdti
etc.) loses all value, for Yaska appears to have hit upon preciselv the
folk-et\unologA’ which changed kasydpa- to kacclmpa-.
But in the main Dr. Ghosh's e.xposition is thorough and informative
and his judgment sober. He has added to our knowledge of the formation
of Sanskrit words.
Rgveda-Samhit.I. Part I. Edited by S.vtischaxura Seal and
SiTANATH Pradhax. Indian Research In.stitute Publications :
Yedic Series. x 10, pp. Anii, 102. 22. Calcutta. 1933. 2.5. 6d.
The difficulty and expense of obtaining a weU-printed text of the
Rgvmda, either in India or elsewhere, has long been felt. A moderately
216
REVIEWS OP BOOKS
priced, accurately printed text alone would have been very welcome.
The plan of the editors of this publication is however far more grandiose.
When completed, it is to contain the accented text, the Padapatha,
Sayana's Commentary with extracts from other commentators such
as Skandasvamin. Yehkatamadhava. Gunavisnu. Anandatirtha, etc.,
together with variant reading.s and e.xplauatory and critical notes in
Sanskrit ; an English translation (in the light of Sayana's Commentary)
and notes containing the interpretations of Western scholars ; and
finally two other translations in Hindi and Bengali.
This first part contains an Introduction in Sanskrit, the Introduction
to Sayana’s Commentary, and the first two hjmins of the Eg\'eda, with
their English translation ; but no Hindi or Bengali translation. There
is then not much on which to form a judgment. Of the text and com-
mentaries there is nothing to be said except that there is here some hitherto
unpublished commentary and that unforttmately misprints are by no
means absent. Every effort should be made to avoid these in future.
The English translation, by Sitanath Pradhan, with its notes, is praise-
worthy, but it suffers from the drawback that it is based on Sayapa.
The translator has made wide, though not exhaustive, use of Western
scholarship, and— perhaps one of the best features — extensive collection
of parallel passages. It is unfortimate that in his romanized text he has
not put in accent marks, thus necessitating constant turning back to the
devanagarl text. This could be corrected in subsequent parts. The
rules of the Institute prescribe that one fascicule shall be published everv
month. This one. published in August 1933, is all that we have received.
Whether or not Rule 10, “ The whole Rgveda-samhita will be published
in five years, if not earlier will be carried out, we cannot say.
ApastambIyadhar.vasutram : Apastamba's Aphorisms on the Sacred
Law of the Hindus. Edited in the original Sanskrit, w'ith critical
notes, variant readings from Hiranyakesi-Dharmasutras, an alpha-
betical index of sutras. and word-index, together with extracts from
Haradatta's Commentary, the Ujjvala. By G. Buhler. 3rd ed.
7 X 10, pp. xii, 280. Bombay Sanskrit Series, 1932. Rs. 3.
Hr. H. G. Shastri has performed a serGce to Sanskrit studies by
seeing through the press this third edition of Biihler's well-known
Apastambiyadharmsutm.
LIKGUISTICA
217
The Unahisuteas with the Vrtti of Svetava^'avasds. Edited by
T. R. Chintahan'i. Madras University Sanskrit Series Xo. 7.
6^ X pp. xiii, 236, 46. University of Madras, 1933. 6s.
This is the first part of a projected series of the Unadisutras in various
recensions. The most popular of the Unadisutras is the one commented
upon by Ujjvaladatta and Bhattoji Diksita. This is the one, belonging
to the system of Panini, on which the commentary of Svetavanavasm
has been here edited by ilr. Chintamani. The Introduction contains
a note on the author of the commentary, but discussion on the disputed
question of the authorship of the text itself and of Unadi literature in
general has been left to the last part of the series. The editing of the
text, based on three MSS., has been competently carried out ; and the
usefulness of the volume is greatly enhanced by the indexes of sutras
and words. The University of Madras is to be congratulated on the
continuation of its Sanskrit Series.
Nandike^vara’s Abhinaya-darp.anam. Edited by Manomohan
Ghosh. 6 x 9J, pp. Ixxii, 66, 55. Calcutta Sanskrit Series, 1934.
Conventionalized gesture plays a great part in the dancing and
dramatic representations of India and the further East. We welcome
therefore this excellent edition of the Abhinaya-darpana of Xandikesvara
prepared by Mr. Manomohan Ghosh. In addition to the text, provided
with an apparatus criticus (based on two complete and three fragmentary
MSS.), Mr. Ghosh has given an informative introduction, an English
translation, a selected glossary, and a good index.
The Padyavali of Eupa Gosvamin. Edited by Sushil Kum.ar De
Dacca University Oriental Publications Series, Xo. 3. 6| X 10,
pp. cxliv, 296. University of Dacca, 1934. Rs. 7.8.
The text here ably edited by one whose name is so well known in the
field of Sanskrit Poetics is a Vaisnava anthology in Sanskrit compiled
by Rupa Gosvamin, who flourished in Bengal in the fifteenth century.
In addition to the text with a full apparatus criticus based on sixteen
MSS.. Dr. De. in a long and most interesting Introduction of over 100
pages, has dealt with the life and works of the author and the development
of the Caitanya movement and the Bhakti-cult in mediaeval Bengal.
He further greatly enriches the literary and historical value of the
book by notes on the authors of the various poems included in the
anthology". Admirable indexes add to the accessibility’ of the information.
The whole is a worthy expression of Dr. De's ripe scholarship.
218
EEVIEWS OF BOOKS
Ox THE Moderx Ixdo-Aryax Yerxaculass. By G. A. Griersox.
Eeprinted from the Indian Antiquary, vols. Ix, Ixi, Ixii (1931-3).
85 X lOj, pp. 186. British India Press, Bombay, 1934. £1 5s.
It is some consolation to those who deplore the decease of the I
Antiquary, that before it came to an end the editor was able to publish
this contribution of Sir George Grierson to the studies in which he is the
doyen. These pages contam part of the material which he had collected
over many years for that volume of the Grundriss der I ndo-AriscJien
Philologie und Altertuni-skiinde which was to deal with the modern Indo-
Aryan languages. Although to the profound regret of all Sir George has
felt himself unable to complete this work, it would have been even more
regrettable if the great mass of material he had gathered and all the ripe
wisdom that lay behind it had not seen the light of day. For his generosity
in putting it at the disposal of his successor for the volume in the
Grundriss could not in any way have compensated for the loss of the
presentation of it by himself. All the more then we may rejoice that
this invaluable contribution to Indo-Aryan studies was not only pub-
lished in the Indian Antiquary, but has now been issued as a separate
volume. This contains two introductory chapters (General Survey and
Historical) which have already appeared, in a preliminary form, in
Vol. I of this Bulletin. As this volume is now out of print, it is all the
more valuable to have these chapters, with numerous additions, printed
again here. Chapter III (Pronunciation and Alphabets) is provided
with seven plates illustrating the northern alphabets. Chapter IV
(pp. 69-186) contains the main body of the work, the Phonology, in which
Sir George sets out the main changes that have taken place between
Sanskrit and the modern languages. In the midst of this great mass of
detail, there are naturall}' points on which not all. including the present
writer, would agree with Sir George's conclusions ; but his unrivalled
knowledge of Indian languages, his strong intuitivm sense and the long
years of patient toil have collected here a wealth of material which will
serve as a mine for many future scholars.
Pahud.vdoha of R.ajia.si.mha Hi'xr. Edited by Hiralal Jaix.
Karanja Jaina Series. Vol. III. 4f x 7, pp. 7, 136. Jaina
Publication Society. Karanja. 1933.
KArakamdacaeiu of Kaxakamara Mrxi. Edited by Hiralal .Iaix.
Karanja Jaina Series, Vol. IV. 6i X 9i, pp. 50, 284. Jaina
Publication Society, Karanja, 1934. Rs. 6.
LINGUISTICA
219
Jayadhavala Tika of VIeasena. Edited by Htrat.al Jain. x 9i,
pp. 19. Saraswati Press, Umraoti, 1934.
In editing these three Apabhramsa Jaina texts Professor Hiralal
Jain agains throws a debt of gratitude on all students of Apabhrarim
and of Jainism for his inexhaustible activity as an editor. The first
of these texts has a long Introduction in Hindi discussing Jaina mysticism,
the date of the composition, and the relation of the terms deslbhasd and
ajpabhramsa. Opposite the text itself is a Hindi translation. The
second volume, which contains the life of the Pratyekabuddha Kara-
kanda, has a short Enghsh introduction, an English summary of the
poem, the text itself with critical apparatus based on five MSS., and an
English translation. Both have that most excellent feature to which
Hiralal Jain has now accustomed us, a complete index of all words with
references to the text and Sanskrit equivalents or translations. The
much shorter Jayadhavala Tlka is provided with both a Sanskrit chdyd
and a Hindi translation.
Krsnajanma of Manbodh. Edited by Sri Ume^a Mi^ra. x 9^,
pp. 9, 56, 7. Yidyapati Press, Laheria Sarai, 1934.
In 1882 Sir George Grierson published in the Jonrn. As. Soc. Bengal
Yol. LI the text and in 1884 Yol. LIII Spec. No., an English translation
of ten chapters of the Haribans of the Maithill poet Manbodh. Now
Sri Umesa Misra has given us under the title Krsnajanma the text
of the complete eighteen chapters, based on a comparison of five MSS.
the variant readings of which are given in the footnotes. The text of
the first ten chapters differs little from Grierson's except for the addition
of three lines in Chap. IX and one each in Chaps. II-Y, YII, YIII.
In the Introduction, Muitten in Maithdi, the author makes some remarks
on the Jlaithill language and its relation to Sanskrit. He does not go
verv deeply or completely into the subject, and lias evidently no special
acquaintance Avith Indo- Aryan philology as a whole — the remark, e.g.,
that Sanskrit final -a is changed to -i in Apabhrarhsa and Maithill as
eA-idenced by Skt. prstlm-. IMth. fiUd AAould haA'e been corrected by
reference to lYcp. Diet. s.a'. pith Avhere the forms in -I are shoAvn to be
deriA^ed from Skt. pMi-. But it is the text Avhich matters, and it is A-erv
good to haA'e that complete and critically edited. One thing more:
how immenselv the linguistic A’alue of the book Avould haA'e been increased
by the addition of a complete index of AA'ords.
220
REVIEWS OF BOOKS
A Phonology of Panjabi as spoken aboht Ludhiana and a Ludhiani
Phonetic Reader. Bv Banarsi Das Jain. Panjab Lni\ersity
Oriental Publications. Xo. 12. x 9k pp. vii. 226. University
of the Panjab, Lihore, 193J.
Dr. Banar.si Dai .Jain ha.s for many years been connected with the
Panjab University's undertaking to produce a Panjabi Dictionary.
It is earnestlv to be hoped that this plan, for which so much work has
been done bv Dr. Jain, will early see completion. In the meantime the
Universitv is to be congratulated on publishing this work, which had
its origin in a thesis for the L'niversity of London. Dr. Jain has pro-
duced a really useful sketch of both the phonetics and the phonology
of the Ludhiana dialect of Panjabi. The importance of Panjabi and
especially Eastern Panjalii in the linguistic history of India is heightened
by its close connection with, and influence upon, the Hindustani which
is already the lingua franca of Xorthern India.
A considerable ad\'ance is made by the author in the vexed question
of accent and its effect on 3IodIA. phonolog}'. He more closely defines
the conditions in which a Sanskrit syllable becomes the .stressed or
prominent syllable of a Panjabi word : in effect the first syllable from
the end of the word containing a long vowel, or, in the absence of a long
vowel, the first heayi' syllable (i.e. short vowel followed by two or more
consonants) from the end of the word, remains a long or heavy syllable
to the present day, whereas the other syllables of the word are liable
to various shortenings.
A few points. While the long 7 of jlbh (jihvd), rluhd {arista-), niajilh
{manjisthd) may be explained as borrowing from the Hindi type, the
same cannot hold for Jatld ' did ', in which, as I have suggested for SindhI,
the 7 may be due to contamination with a form *h'ia- < *kiaa-. The -tt-
of klttd is not necessarily an analogical extension of the -tt- in the tvpe
siittci {suptd-) but is just po.ssibly a descendant of *kitra- < Irtd- attested
indirectly in Asoka Shah, kitra- and perhaps in Skt. krtrima- and kidlitna-. ^
For mater (p. 39) the Skt. cannot be *nmlritara- for vvLich tvye there is
no authority, but ^mdtrtard (cf. mdt'rtama-, and Lat. mdtertera ‘ aunt ')
> *mdtritard > *mdttiara. The development of r after plosives to ri
{ru) is known for Asoka and the NW. Prakrits ; it was probablv' of wider
extent as an intermediary for Pkt. i, u than appears, for traces can be
seen when, as here, r follows an internal plosive : .so probably Pkt. appiUtha-
' tt must be confessed however that .Si. kito, with t not tr, and Lah, klttd, with tt not (r,
are against such a theory, unless kitra- i.s to be read as kirla- (rf. kitri = Skt. klrti-),
and the form in question here he *kirta-, not *kitra-.
LINGUISTICA
221
< aprsta- through *apnista- (cf. Ksh. prichiin ' to ask ') rather than
< a- + puttha- with analogical doubling of initial p-.
It is to be hoped that the author will continue his studies of Panjabi
dialectology, an excellent example of which was recently seen in this
Bulletin.
SRiMADBHAGAVADGITA KA DoGRTbHASAN’UVAD : ThE BhAGAVADGITA
WUTH THE TEXT IX SaXSKRIT AND TRANSLATION IN DoGRI. By
Gauri Shankar. 4| x 5^, pp. vii, 204. Lahore, 1935.
3Ir. Gauri Shankar, who has already published interesting information
concerning the Dogri language in Indian Linguistics, has now, in the
words of Dr. A. C. Woolner's foreword. given an example of the
literary capacity of his dialect by making a translation of the Bhaga-
vadglta."’ Dogri is spoken by about half a million people, and records
of it are urgently required. Since such a translation as this necessarily
contains a large number of learned words, we may echo the wish of the
Vice-Chancellor of Lahore University “ that it will be followed by another
work giving us an idea of the idiomatic village speech and the daily
life of Dugar ”.
Selections from Classical Gujarati Literature, Vol. 11. By
I. J. S. Taraporewala. 9 x 6|, pp. 10, 684. The University of
Calcutta, 1930.
These selections from Gujarati literature, of which the first volume
appeared in 1925, are intended to do for Gujarati what Dineschandra
Sen's Selections have done for Bengali. Both these series owe then-
inspiration to the late Sir Asutosh IMukerjee. Dr. Taraporewala, at
one time Professor of Comparative Philology in the University of Calcutta,
has done much to arouse enthusiasm for the study of the Modern Indo-
Aryan languages and their history. The selections in this volume cover
the sixteenth and seventeenth centuries. It will be of greater value
for the student of literature (for whom it is in the first instance intended)
than for the Imguist. since some at least of the selection.s. although the
editor does not give the sources, appear to rest on considerably modernized
texts.
A Primer of the 5Iarathi Language : for the use of Ai)ult.s.
Bv A. D.irby. 3r(l edition. 5J X 8^. pp. vi. 221. Tatva-xfivechaka
Press. Bombay. 1933.
This primer of IMarathi deserves notice here not simply as a text-
book for learning that language, but as a ivork in which the author has
222
REVIEWS OF BOOKS
deserted the traditional grammatical description based either on Sanskrit
or on Latin grammar and has described Marathi grammar pirrely from
the functional point of mew. This is a great advance and is heartily
to be recommended. The description of modern languages, and none
more than the modern Indian languages, has too long been fettered
by the traditional forms and terms invented to describe quite other
types of language. In this respect I should have been glad to see
Mr. Darby go still farther and eliminate altogether the Sanskrit ideas
of case, reser\Tng for Modern Marathi generally the two cases only,
direct and oblique with its various prepositions or postpositions.
The Script of Harapp.\ and Mohenjodaro akd its connection wtth
OTHER Scripts. By G. R. Hcnter. with an Introduction by
S. Langdon. X 9|, pp. 210, plates xxxvii. Kegan Paul,
Trench, Trubner and Co., Ltd., London, 1934.
Mr. Hunter, following his article in JRAS.. has made a much more
extensive study of all the material available, including manv more
seal inscriptions e.xcavated since the material, e.xamined by Smith,
Gadd, and Langdon in Mohenjo-Daro and the Indus Civilization, was
placed at their disposal. The method adopted by ]Mr. Hunter ” in
elucidating the script has been to tabulate every occurrence of each
sign together with those signs whose morphography suggested the
possibility of their being variants ”. It is beyond my competence to
appraise the results of .Mr. Hunter's work : I desire here only to draw
attention to the fact that this work ha.s been done and is available for
the use of other investigators. For whether or not IMr. Hunter's con-
clusions, e.g. as to the meaning and phonetic value of certain smnbols,
or as to this script being the source of Brahml and other forms of writing,
will stand the test of further research or discoverv. there can be no
doubt as to the usefulness of the great mass of carefullv analvsed material
which is collected here.
A Santal Dk tio.vary. By P. 0. Boddevg. 'V'ol. 11, D GH. Vol. Ill,
H-KH. 7 X lOi, pp. .o48, 7.u2. Dybwad, Oslo, 1934-5.
The Xorske Yiden.skap.s-.A.kademi i Oslo continues the publication of
P. 0. Bodding's monumental dictionary of the Santal language, of
which the first volume has already been reviewed in this Bulletin All
that was saitl there, both of the high linguistic importance of this
work and in praise of its performance, may he repeated here. .Ls again
appears, the very extensive invasion of the language by Aryan elements
LINGCTSTICA
223
indicates how necessary it is to record these interesting siuA-ivals of
a pre-Aryan India before they disappear from the scene. That the
life-work of so devoted a lover and student of the Santal people should
he published through the munificence of a foreign Academy is a matter
of rejoicing for scientists but of reproach for the Governments and
institutions of India whose prerogative it should be to record the languages
of the peoples under their rule or within the sphere of their mfluence.
The Beahui Language. By Dents Beay. Part II, The Brahul
Problem. Part III, Etymological Vocabulary. x 9i, pp. ii,
313. Manager of Pubhcations, Delhi, 1934. 13s. (kl.
Of the two great language-famihes in India, the Indo-Aryan and the
Dravidian, the history of the former is almost infinitely better known
than that of the latter. For with Indo-Aryan not only have we docu-
ments composed in a language which was not far removed from the
common original, but a vast amount of work has been done from the
linguistic point of view on the rich material available from the ^gv'eda
down to the spoken languages of to-day. True, most of that work has
been directed towards the more ancient documents, hut of late years
Indologists, following the pioneer work of Beanies and Hoernle, stim-
ulated afresh by the carrying through of Grierson's gigantic task in the
Linguistic Survey, have begun the scientific study of the history of
the modern languages. How far that work has proceeded is shown by the
appearance of Bloch's L'hido-aryen da Veda aax temps moderaes
reviewed above. In this, whatever may still remain doubtful in detail,
are disclosed, once and for all, the main lines of the Imguistic development
of the Indo-Aryan languages. Far otherwise is it with Dravidian.
Not only have we no record of the original language, but since the time
of Caldwell, who pubhshed his Comparath'e Gramnmr of the Dravidian
Languages in 1856, long before the Junggrammatiker had revolutionized
linguistic science, no trauied linguists have given more than a fleeting
glance to the problems of DraNudian philology. Not even have the
individual histories of the four great literary languages with their com-
parativelv ancient documents yet been aderpiately studied. But in
addition to these there are the uncultivated DraN-idiau dialects of the
north, for the most part isolated little islands doomed before long to be
swamped by the surrounding .seas of Indo-Aryan. For these, presenting
invaluable material for the comparative grammarian, the most urgent
immediate need is their full description, before they finally disappear.
All the more welcome then is the completion of Sir Denys Bray's
224
REVIEWS OF BOOKS
work on the Brahiil language. For Brahui. though not in such danger
of early decease as some of the others, is especially interesting to the
comparativust owing to its remote situation in Baluchistan and its earlier
separation from other Draiudian languages.
In Part II (pp. 1—43) Bray gives what is known of the history of the
Brahuis and discusses the various theories that have been put forward
as to their racial and linguistic origins. Here his sober judgment finds
few certainties except the fact of the Dra%'idian origin of their language.
He permits himself one speculation, based as he confesses on little sa^ e
contiguity, namely that the Brahuis may be the remnants, or at least
the linguistic descendants, of a pre-Aryan Dravidian population whose
civilization is represented for us now by the ruins of Harappa and
Mohenjo-Daro.
The value of Part III cannot be overestimated. For we have here
an excellent vocabulary of the language, in which the words are richly
illustrated by examples of the contexts in which they are used. How
far it may be properly called an etymological vocabulary is a little doubtful,
for the author has for the most part contented himself with very general
comparisons and indications. In the case of the Dra%fidian content
that was inevitable, for we have as yet established no precise corre-
spondences between the Dra\udian languages. But for the innumerable
borrowings from Baluchi, Lahnda (Jatkl), and Sindhi further precision
might have been attained. Usually Bray has quoted forms from several
languages without indicating which is the probable source of borrowing.
In some cases the form indicates a Lahnda rather than a Sindhi
source : so, with double consonants where Sindhi has single : —
chalckl ‘ taste chaUing ‘ to lick ', gudii ‘ kite ’, kukkur ‘ cock ’ (as
shown in the example) from Lah. kukkur m., not kukkir f. or Si. kukir* f..
lagging ' to meet ’, pinnl ‘ calf of leg ', pnssi ‘damping ’ ; or with
dental before r : — drah ‘ a partic. wild grass " (Lah. drabh, but Si.
drah}i'‘). triJidn ‘ 3 year old camel ’ (Lah. trihdn, but Si. trihdrp‘).
On the other hand the double consonant may be due to borrowing
from an older stage of Sindhi before the simplification of double
consonants in the case of words not attested with a double consonant
in Lah. ; — baceing ‘ to escape ’, budding ‘ to sink bugfp ‘ fried
jann ' bridal procession ‘ (Lah. janj. Si. jan'^). lattur ‘ trampling '.
Similarly a final -an appears to represent the older stage of Sindhi -o
in bnnnan ‘ embankment '. butnu ‘ bush '. tnruu ‘ saddle-felt and
perhaps manjav ' hurdle made of branches ’ (contrast the more
modern borrowing of gnndo from Si. gundho).
LINGUISTICA
225
Doubtless the change of -I- to -r- in Sindhi is comparatively recent :
we accordingly find forms with -I- where modern Sindhi has -r- fairly
common, sometimes with other distinctive Sindhi characteristics, as
in dukJcal ‘ dearth koddl ' mattock ’ (Lah. kddal, Si. kwjar', cf.
koddryo ‘ worker with k.'). The case of sulung ' tunnel for which
Bray quotes Lah. stdioty beside Si. surihgh'^, is more difficult. Is
-I- a Brahui innovation, as apparently not uncommonly in the Nushki
dialect, or is it borrowed from some lA. dialect with ? So jal
‘ torrent-bed from which B. rightly dissociates Skt. jala-, etc., may
be compared with Skt. lex. jhara- ‘ cascade survivdng in Bg. Hi.
jharl f. ■ waterfall Gu^.jharo m. ‘ fountain Mar. j'/iara m. ‘ spring ’ ;
-I- forms in the compound Pkt. ttjjJialia- ‘ thrown up Hi. ujjJialna
■ to flow from one vessel to another Or. ujhidaibd ‘ to clean out a
tank ’.
The change of v~ to b- is perhaps recent, and need not be taken as
an isogloss running through other Dravidian languages. At least
lA. V- appears as b- in a number of words : bdhot ‘ refuge ' which B.
derives from Sindhi ‘ protection ’ + ■' shelter bitting ‘ to
pick up ’ for which B. quotes Hi. binna, better the older blnnd < Skt.
vicinuti : in the West and North-west I have only found this in Guj.
and Kash. {Nep. Diet. s.v. binnu). bashsham ’ summer rains ’ (also
in Baluchi) is perhaps a loan from Mid.I.k. < Skt. rarttdh with
Brahui suflix -{d)m.
Morgenstierne (NTS. ^’) has shown that Baluchi loanwords in
Brahui often show an older form of the language. The same is true
of lA. as hinted by B. in Part II. bahu ‘ armful ’ with the final -u
of the Skt. stem which was replaced by -d in Si. etc. bd/t" ‘ arm
while in Lah. bd/tii has the meaning ‘shoulder of horse’. For Skt.
vrkkdii Pa. and Sgh. have e.xpected forms with va- : elsewhere in India
including Dardic we find only bru-, bu- (see Nep. Diet. s.v. boknu) :
Brahui bik ‘ kidney ’ points to a Mid.IA. *i'{r)ikka-. We have seen
Mid.IA. s in baslishdm, perhaps also in kashshing ‘ to pull out ' <
Mid.IA. *ka^s- < Skt. kdmiti. though tiis/ish ‘ bran ’ with double sJish
indicates the possibility of kashsh- being a loan from Pers. kas- ; and
kamdsJi ‘ greyhaired for which B. tentatively suggests Skt. hdmdm-.
Skt. s in shank ‘ twig ', Skt. sahkn- which survives in Khowar sogg
‘ twig ’ (Morgenstierne) and in other Dardic dialects ; shevdl ' waterplant
vallisneria octandra ’, Skt. saivdla- (Nep. Diet. s.v. shrdli) : shik
‘ pointed ’ (see below) : kdshitm ‘ saccharum ciliare '. Skt. kdsa- (Nep.
Diet. s.v. kds). Skt. -mb-, -nd- which became -(ni)m-, -(n)n- in the
VOL. Vin. PART 1. 15
226
EEVIEWS OF BOOKS
neighbouring lA. languages, are maintained in ildmb ’ gipsy blaoksmitli
kanihar 'spotted': land 'penis' (ISep. Dirt. s.v. Idra). pindhtq 'to
beg Skt. nl in nnrdhit/ ' to groan '. 8kt. ndnlnti (or from Iranian
*nard- in Pashtu nor-. Morgenstierne Ef. Voc. PasJitv. p. 53). Skt.
intervocalic unvoiced plosives appear to be preserved in some words,
indicating very early liorrowing ; .v/n'A' ' pointed '. Skt. i'lkhd ; I'dfi
' hornless ', lat ' .shaven ', Skt. k f/ta- ; pntaii ' .smallpot '. Skt. gpliotn- ;
chihit ' unripe melon ' beside Si. cib/iir’‘ ' cucuniis pubescens ' < Skt.
cirbhata-, cirbhita-. dranna ' the wild bu.sh artemisia ' points to
earlier *dni)iiua- like Khowar drdn ' worniwood ' (Lorimer), while
Skt. damana- or *dra»iana- appears in all the other lA. languages
(Nep. Did. s.v. dainand). This makes more plau.sible the comparison of
chuh ■’ ravenous ' with Khowar chii ' hunger ' (LSI. ; for chttl '<) < Skt.
hsudhci which appears in Pa. as Ihudd. in Pkt. as chiJid as well as
khnJid, while Pj. has only khoh. pihl ' goat dysentery with ulcerated
liver , if from the family of Skt. plVuin-, agrees with West Paharl
Bhalesi tki 'spleen (Siddheshwar Varma) and IMar. pihu: elsewhere
in Mod. lA. pi- > pil- (Eep. Did. s.v. phitjo). Skt. IdnyaJa- '' plough ’
attested in Bg. Jdiinl, nuiml, Magahi Jdfjal, Bih. nnqann.^ Hi.
nagal, Sgh. aagala. aangal and in the X.W. in Bhadrawahi
ndngal, appears in Brahui as Uoujcir and Baluchi nangdr. pappus
‘ lung ’ is more like Waigeli papns, Kati p<->pyus. Pa.shai papu (cf. Skt.
jjhupphusa-) than Lah. phipphur. dfoltur ’ greed, de.sire to eat in
pregnancy', as derived by B., .supports the a.s.sumption that Skt.
doliadu- ■ morbid desires of pregnancy ' (Yajnavalkya) is a Mid. lA.
form of daurhrda- (only lex. in this .sense), dauhrda- (Susruta).
Even if Brahui has borrowed widely from lA.. the previous
ascendancy of a Brahui confederation would account for manv words
being borrowed from it by it.s neighI)ours ; and in hict we find here
a great many words common to Brahui. Baluchi, Lahnda and Sindhi,
which do not seem to have repre.sentatives in other lA. lanuuagos.
In the present state of our knowh>dge it is often impos,' ible to
say whether a Brahui word of Dravidian origin has de.-eendcd directlv
or has been re-borrowed from lA. Is kunml ‘ curl ' di ■ ei Ir eonnected
with Tam. etc. kurul. or a loan from .some NIV. form < ^ A, / do/u- I
bel ‘ half a day or night . one of two meals, divi.sii •
with apparently the primary meaning of ' di\-i.sini' n , ; h-spite
B. who objects on .score of meaning, be derived from f i d',' . ’ which
the earlier meaning is not ‘ time ' but ' limit ' ; but it n rv ho be
real Dravidian belonging to the family of Kan. Idr ' . m u-; tion '
LINGUISTICA
227
from which Skt. veld was a loan as suggested bv Kittel. Kan. Diet.,
p. xlii. On the other hand, hatting ' to pound ’ wliich B. appears to
connect directly with Tam. etc. huttu. is certainly a loan from lA.
hutt- (itself probably borrowed from Dra\ddian), since the normal
correspondence of Tam. Kan. h- is Braliui W.
It will be seen what great interest this work will have first and fore-
most for the student of Dravidian, but also for those concerned with
the vocabulary of Indo-Aryan and of the neighbouring Iranian
languages. We must indeed be thankful that the Government of
India undertook the publication, but we may regret that so important
a work had to be printed at the Government Press, with its ugly t\'pe,
its small resources (only one fount is used throughout the dictionary
for Brahui, English, and all other languages alike), its bad alignment,
its rather frequent misprints, and its execrable binding. The result
of so many years of devoted work deserved a better habitation.
Ixdo-Uralisches Sprachout. Uppsala Universitets Arsskrift 1934.
Filosofi, Sprakvetenskap och historiska Vetenskaper 1. Yon
Bjorn CoLLixDER. pp. 116. Uppsala, 1934. Price: Swedish
Kronor 3.50.
In this excellently printed and reasonably priced book the question
of the possible relationship between the Indo-European and Uralian
(i.e. Finno-Ugrian plus Samoyede) families of languages is discussed.
The evidence is clearly presented and the discussion is exhaustive
and concise. 3Iuch has been written on this subject, but
Professor Collinder's work will be welcome as an exceptionally clear
and methodologically rigorous pronouncement on the problem, more
particularly in these days when all too many rash suggestions as to
distant connections of both Indo-European and Uralian are being made
(cf. the recent attempts to show that Indo-European and Corean, and
that Uralian and Eskimo are cognate). Further, to quote Collinder's
own words (p. 16) : “ Die vorliegende untersuchung bringt sowohl
in morphologischer wie in lexicalischer hinsicht bedeutend mehr
beweisstoff als die vorgiinger. Und zugleich ist jedenfalls der
lexicalische stoff strenger gesichtet, als es bei den vorgangern der
fall war."
The question discussed by Collinder is, a priori, of the very greatest
importance. Among other things, if it can be shown either that IndE.
228
REVIEWS OF BOOKS
and Ural, are related or that the Pr.IndE. and Pr.Ural. speech-
communities were in contact, the much-vexed problem of the “ indo-
germanische Urheimat ■ ■ is all but solved, since the Pinno-Ugrian
urheimat was certainly near the Volga. It is remarkable that the
Fimio-Ugrian aspect of this problem has been so much neglected
by Indo-European philologists.
In' an introduction (§§ 1-22) a survey of the literature of the subject
is given. Many of the best -known names of Finno-Ugrian philology
(e.g. Vilhelm Thomsen, Anderson, Budenz. Munkacsi, Wiklund,
Setiilii. Paasonen, Ojansuu, Skbld, Jacobsohn, Lewy) and a few well
known in the Indo-European field (e.g. Sweet, Holger Pedersen,
Joseph Schmidt) find mention here. Further the problem raised by
the fact that there are no correspondences between the IndE. and
Ural, numerals — which, on the hj'pothesis of ultimate relationship,
might at first sight be expected — ^is dealt with.
Then follows a list of Pr.IndE. and Pr.Ural, correspondences, with
sections on morphology and Ic.xicography (in the latter the pronouns
are di.scussed in a .separate sub-section). The following correspondences
are given ^ ; —
Morphology
Inflexion
DECLENSION
Plural
§ 23. PI. suff. -i- : Finn. pi. stem linnui- {<lintu “bird”)
Tavgy abl. pi. kulagita : abl. sg. kuh^ata (< kula “ raven ”) co IndE.
*lo-i (= Goth. y>ai).
Case-endings
§ 26. Acc. sg. : Mari ^iiSom {< ^iit “water’’) Tavgy kulam
(< kula “ raven ”) Lat. lupum.
§ 27. Abl.; Mord. abl. sg. tolgada, tolgado {<tolga “feather’’)
takada adv. " from behind ” co Oscan sakarakl lid “ sacello
§§ 29-35. Casus rectus ’’ in t (+ vowel) : Mari tuhi) “ he ”
(< tti that one ”) Scl. kud, kudo “ who ” (cf. Finn, ku-ka “ who ? ”)
CO IndE. *to-d (= Skt. tdt ; cf. also Goth, pnt-a),
§§ 36-41. Casus obliquus ’’ in -n-. Here Collinder discusses
the problem afforded by the appearance of an n in certain positions
1 To save space only one Finno-Ujirian, one .Sanioytxlc, and one IndE. form are
given from each sot of cognates. The .Samoyeilc ciialeit.s are referred to as Yurak
Tavgy. Pel. ( = Ostyak-Samoyede. Huss. ,cVlvp.hju Kam. = Kamass. Motor; thJ
Finno-Ugrian languages by the names .suggested as standard in Leeds Studies in
English anA Kindred Languages n'i, 58-9.
INDO-URALISCHES SPEACHGUT
229
in the “ conjugation " of nouns with personal suffixes. Paasonen and
Mark assume that Mansi, in which the n indicates that the object
possessed is in the plural (kwol “ hou.se ” : kwolem " mv house":
kicolanem “ my houses best preserves the Pr.Ural. state of affairs
but Collinder (following AViklund) shows that in Pr.Ural. this n
occurred in the oblique cases of the singular as well as in the plural
(cf. Erza-3Iord. noni. sg. kudom " my house " : iness.sg. kiuloson
“ in my house " ; noni. pi. kudon “ my houses " Yurak Imnba “ snow-
shoe ” ; abl. sg. lambahad “ from a snowshoe " : nom. acc. sg. Jambau
“ my snowshoe ” : abl. sg. lambahculan “ from my snowshoe " : nom.
pi. lamhin “ my snowshoes He then goes on to connect this Ural.
n with the n of the oblique case of the IndE. heteroclitic declension
(cf ILat. femur : feminis) which he believes to represent a very early
IndE. state of affairs.
CONJUGATION
Tense
§§ 46-8. s-preterite : Mansi 1 sg. pret. minssm (< min- “ to go ”)
Tavgy 1 sg. “ 2nd tense ” tumtusuam (: 1 sg. “ 1st tense ” tumta'am)
“ make fire ” oo IndE. s-aorist (as in Gk. eSet^a < BelKvvfii).
Personal Endings
§§ 49-50. 1 sg. ; Pr.Ural. -tn (Hung, alszom “ I sleep ”, Yurak
madam “ I cut ”) co IndE. -?m, -m (Skt. dsmi “ I am ”, dhharam
“ I carried ”).
1 pi. : Pr.U'ral. -mek, -mak (Lapp, nmndim-ek “ we went ”, Tavgy
mat'amii we cut ") co IndE. -m + ? {Eat. ferimus).
2 pi. : Pr. Ural, -tek, -tak (Hung, hallotok “ you hear ”, Yurak
rnadada' “you cut ”) CO IndE. -te (Gk. ^epere).
Suffixes
§§ 51-72. Here follows a list of correspondences in the suffixes.
Lack of space renders it impossible to give a summary of this section,
which, since both the semantic and the morphological aspects are
involved, would have to be full to be of value. A large number of such
correspondences are given.
Lexicogr.4phy
Pronouns
{a) Personal Pronouns {and {b) Possessive Suffixes)
§ 74. 1 sg. : (a) Mord. mon, Sel. man “ I ” ; {b) Hung, -m {nevem
“ my name ”), Tavgy -rna co Eng. me.
230
KEVIEWS OF BOOKS
§ 75. 2 sg. : {a) Mord. ton, Sel. tan ''' thou ; {h) Mord. -t {kudot
“ thv house ”) ^ Lat. te.
§ 76. 3 sg. : (a) Mord. son “ he, she ; (b) Komi -s {karys “ his
town ’ ) ^ (a) Gk. ol (P) Goth, si {y) Lat. se (8) Goth, sa, so.
Demonstrative
§ 77. («) Finn. noni. sg. tanui, stem W-, Sel. tarn this ” ; {b) Mord.
to-, Sel. to “ that co Gk. to.
§ 78. (fl) Finn. nom. acc. pi. nCiniat, stem nil- " these ■’ ; nuo.
niiot ■■ those ", Sel. na “ that one " ; (b) Finn, niin, niiin, noin " thus ’,
Sel. nild'el thus, .such " oo Skt. na-nd " thus and thus ".
§ 79. Hung, e-z Kain. Ide " this " &o Goth. is.
§ 80. Hung, a-z Lith. n-ns ‘‘ that ".
Relative
§ 82. Finn, stem jo- Skt. yds “ who
Interrogative
§ 83. (A) Hung, ki co Lat. guts “ who ? ’’
(B) Finn, stem ku- " who ? ", Tavgy kua “ which ? ” co (a) Goth.
kieas (b) Albanian kit “ where ? ’’
Alphabetical List of Nouns, Verbs, and Particles
§ 84. Finn, asn- " to dwell ”, Yurak -gesy “ tent, ^^llage of tents ”
CO Skt. nsmi “ I am ", asta- “ dwelling
§ 85. Hung, htir “gut, cord", Yurak hurku “(sinew-) cord”
CO Gk. yopSg “ gut, cord '.
§ 86. Hung. kap-“ to seize, get’’, Kam. kuiva-r-lim “to seize”
CO Lat. capio.
§ 87. Finn, keri “birch-bark”, Sel. keram “to peel, skin”
CO O.Bulg. skora, kora “ bark ".
§ 88. Finn, -ki “ also " Kam. -ko “ and, also ” co Lat. -gue.
§ 90. Mord. knnda- " to seize ” co Lat. prae-hendere.
§ 92. Vatja kuraz “ knife ” Motor kura, kitro “ knife ” co OE.
heorii.
§ 94. Finn, lante- “low", Udmurt hid “field”, Yurak laindo
“low” CO OPru-s.s. lindan “valley”, Eng. land.
§ 95. Mari mane-, Yurak man- “ to say co German mahnen.
§ 96. Hung, nm-, Yurak mdsa- “ to wash ” co Lith. mazgoti
“ to wa.sh ”.
§ 97. Finn, murta- “ to break ”, Sel. inornam “ to smash ” co Skt.
mrnCiti “ grinds, smashes
INDO-URALISCHES SPEACHGUT
231
§ 98. Finn, myy- “ to sell ”, Yurak mi- “ to give, sell ” co Lat.
munus.
§ 99. Finn, nato ” sister-in-law ”, Yurak nado wife's younger
brother ” co Lett, zndts “son-in-law, brother-in-law”.
§ 100. Hung. He “ not ", Yurak tii “ not he ” co Goth. ni.
§ 101. Finn, nime-, Yurak nim “name” co Eng. name.
§ 102. Erza-Mord. itkso. uJcs " ash, elm ”, Kara, izo ” poplar ’’
CO Eng. ash.
§ 103. Finn, paljo “ much ”, Yurak paju, pal “ thick, close ”
CO Gk. TToXvS.
§ 104. Mari pot “kettle”, Sel. peda, pdta “earthenware pot”
CO Eng. pot.
§ 105. Mord. pele-, Yurak jal/'n “to fear” oo Goth, us-fihna.
§ 106. Khanty pir-dt “last year”, Yurak par-dari “back”
CO Gk. Trip VOL.
§ 107. Lapp, hirrd “around”, Sel. pdr “ ring ” co Skt. pari
“ around ”.
§ 109. Finn, puno- “ to twine Sel. pan-, pun- “ to twine,
braid ” oo Eng. .spin.
§ 110. Finn, snlaea “ salix caprea vel fragilis ” co Lat. .salir.
§ 111. Mord. san, Ta'wgy far] “ sinew ” co Eng. sinew.
§ 112. Finn, tuo- “ to bring, fetch ”, Yurak tu- “ to bring, give ”
<o Lat, do.
§ 113. Mord. red, Motor bu “ water co Eng. water.
§ 114. Finn, retd-, Yurak vada-, vdda- “to pull” co Lith. vedu
“I bring".
§ 115. Finn, viha “hate, poison” co Lat. virus.
§ 116. Finn, vuote- “ year " co Gk. eroy.
§ 118. Komi Lsi?!-, Sel. suhemjun " to decrease ’’ co Gk. drfiOovla
“ abundance ” (< V g'^hben-).
§ 119. Hung, csorog- “ to run ", Sel. sormba “ it flows " co Skt.
ksdrati “flows, vanishes” (< \/ g'^hSer-, \/ g'^Ser-).
§ 121. Norwegian Lapp, cares “coarse, stiff”, Sel. saral “ hard,
tough ” CO German starr.
§ 122. Finn, tarkea “to desire’’ co Gk. arepyeiv “to love”.
In a final chapter (§§ 123-34) these lexicographical correspondences
are studied, chiefly from the phonological point of \dew and the con-
clusion drawn which we may here express in the words Collinder uses
in the preface (p. 4) : “ Ich mochte hier nur die ansicht aussprechen.
232
REVIEWS OF BOOKS
dass von alien aktuellen verwandtschaftshi'pothesen — die ural-
altaische nicht ausgenommen — die indo-uralische gegenwartig
jedenfalls am besten bezeugt ist/’
Theoretically, the answer to the question — is there, or is there not,
sufficient evidence to show that two given languages are related (we
can never, of course, show that they are not related) or that the two
speech-communities in question were in contact in their “ primitive ”
or undivided condition ? — would depend ideally on the solution of a
statistical not a philological problem (cf. Leeds Studies n, 7-13), viz. is the
number of phonematic correspondences between them “ statistically
significant " ? We know that “ the distribution of sound over meaning
is a random one ’, i.e. we know of no reason why, for instance, in
English the first phoneme of the word meaning “ long ” should be
[1] rather than, say, [p]. Hence, when we find similarities between
the phonematic patterns of words of the same or similar meanings in
two languages (as, for instance, between the series Eng. long, lean,
learn, life, love, light : German lang, lehnen, lernen, leben, liebe, licht)
we must conclude one of two things : either (A) that the similarities
are due to pure chance ; or (B) that the languages are either “ related ”
(i.e. were once one and the same language) or that the two speech-
communities in question were in contact in their primitive condition
and that sufficient borrowing — in one or both directions — to account
for the similarities in question took place. The philological problem
has thus been converted into a purely mathematical one for, ideally,
the selection or rejection of Solution A depends merely upon whether
the number of similarities is “statistically significant” or not.
Unfortunately this simple method of solution must at present remain
an ideal one for the phonematician.s have not vet provided us with
sufficient data to determine the nature of the random sound-meaning
distribution. But even under pre.sent conditions statistics can be used
in our appreciation of evidence of suggested relationship between
languages.
Thus, turning to Collinder s work, we find one fact that is certainly
statistically significant, viz. (p. 81): “Von etwa zwolf finnisch-
ugrischen pronominalstammen linden sich (mindestens) zehn im indo-
europaischen wieder, und die personalendungen der verba, die ja mit
den personlichen pronomina zu.sammenhangen, weisen eine ahnliche
proportion auf. Von vierzehn finnisch-ugrischen nominalformantia
finden sich zwolf im indoeuropiiischen wieder. . . . Von acht finnisch-
ugrischen kasusendungen haben wir vier im indoeuropiiischen wieder-
INDO-URALISCHES SPRACHGUT
233
gefunden.” This one fact is therefore sufficient to prove alternative B.
In view of this the other correspondences (many of them very striking)
afford strong supporting effidence. though by themselves they could
not afford proof since their number is certainly not statisticallv
significant. Expressing the matter more stronglv than Collinder himself
has done (see above), we may therefore say that the evidence he has
produced does definitely prove one of two things : either (I) IiiclE.
and Ural, are related ; or (II) the Pr.IndE. and Pr.Ural. speech-
communities were in contact and borrowing took place.
Collinder 's book does not, unfortunately, include more than a
passing reference (§ 19) to another important statistical line of approach
to the problem, viz. how many correspondences are there between
Uralian and other arbitrarily selected languages with which there is
no possibility of relationship ? (e.g. one Australian, one North
American, one South American, one Bantu, and one other African
language might have been selected as test-languages). If the number
proved in each case to be very considerably lower than that found
between Uralian and Indo-European we should have further supporting
evidence, though not, of course, proof, for alternative B.
As we have seen, Collinder is in favour of alternative B, but he
only discusses the further question of the choice between hypotheses I
(relationship) and II (early borrowing) in the most cursory manner
(pp. 80-1). While agreeing with his conclusion on this point (relation-
ship rather than early borrowing), the general linguist will be dis-
appointed to find that the complicated question of the a prion
improbability of morphological and pronominal borrowing (which
Collinder, having shown a large number of Indo-Uralian similarities
in the morphology and the pronouns, rightly adduces as evidence
against the hypothesis of early borrowing) is left undiscussed.
One also misses some reference to the much-discussed question ; if
there are very early' IndE. loanwords in Uralian, are they' to
be considered as of Pr.IndE., pre-Ary'an, or pre-Iranian age ?
Methodologically speaking Collinder is no doubt right in excluding
discussion of this question, since he is in favour of the relationship-
hypothesis ; but it would at least have been convenient to have a
summary of some of the chief views on the subject and a bibliography.
As a further point of criticism one would have liked to see some
discussion of the interesting Finn. Mhdekmn, yMehsdn “ 8 ”, “ 9 ”,
the second element of which apparently represents the IndE. form of
Lat. decern, etc. To Collinder ’s list of lexicographical correspondences
234
REVIEWS OF BOOKS
between IndE. and Ural. I should like to add that between the IndE.
root in Plattdeutseh jV/Ae? Eng. ic-vle etc., and the Ural, root in Hung.
jeg, etc., “ ice " which has not, I believe, hitherto been noticed.
Despite these few disadvantages the book and its thesis are of
extreme importance and Professor Collinder is to be congratulated
upon the production of what will undoubtedly prove to be a standard
work. Not the least significant point for the general linguist will be the
fact that the proving of this thesis has shown something that has long
been suspected, namely that general linguistic structure is almost
valueless as an indication of relationship, for there could hardly be
two groups less alike in their general structure than the (so-called)
"synthetic" Indo-European and the ‘'agglutinative" Uralian.
A. S. C, R.
Lappisk Ordbok— Lapp Dxctioxary. By K. Nielsen. Vol. I :
A-F ; pp. Ixvii -f 666 ; 1932. Vol. II ; G-M ; pp. viii -f 718 ;
19C4. Oslo ; Instituttet for Sainmenlignende Kulturforskning :
Publika.sjoner, Serie B, vol. xvii Norwegian Kroner 50
per vol.
V ith these volumes Professor Nielsen's great dictionary is half-
way to completion. It is the largest Lappish dictionary that has
hitherto appeared, and has been eagerly awaited by Finno-Ugrian
philologi.sts for many years. Like Nielsen's other great contribution
to Lappi.sh studies, the Lxrebolc i Lappisk, the dictionarv deals with
Norwegian Lappish and is based on the dialects of Polmak, Karasjok,
and Kautokeino. The dictionary is exhaustive and, both in the number
and arrangement of the entries, reiue.sents an enormous advance on
Friis' dictionary, which has hitherto been the .standard work. Volume I
contains an introduction in which there is a full de.scriptive phonology
of Norwegian Lappish and an e.xplanation of the consonantal ablaut
which is the chief difficulty of the language. The extremely com-
plicated tiansciiption and the .sy.stem of reference.s and abbreviations
used in the work are also carefully explained. The book is of the verv
greatest value and is indispensable to all student.s of Lappi.sh anil
Finno-Ugrian iihilology. It is therefore unfortunate (though in view
of the size of the book doubtle.ss unavoidable) that the price is so hit^h
The reviewer has only two criticisms to offer. The fir.st is one that
might equally be levelled at almo.st all contemporary work on Finno-
Ugrian. It is that the tran.scription used in this work, like that officially
recognized for all Finno-Ugrian languages by the Finno-Ugrian
CHAXDRA-KAXTA ABHIDHAN
235
Society, is obviously phonetic and not plionematic. This has various
disadvantages : it renders it almost impossilde to print the forms
exactly in any publications other than those of the Finno-Ugrian
Society, since nowhere else is tlie enormous nuinl:>er of diacriticized
peculiars necessary for the very complicated transcription available ;
it makes it very difficult (particularly for the beginner) to remember
the forms visually ; and, worst of all, it obscures the essential character
of the phonology of the languages. Perhaps the chief points to be
urged in favour of the transcriptions now employed for Finno-Ugrian
and other languages in the U.S.S.R. are that they arc simple to print
and easy to visualize, and that, inaccurate as they almost certainly are.
they do aim at being phoncmatic rather than phonetic. The second
criticism is one that may seem to come oddly from an English reviewer.
Nielsen's whole work is presented in Xorwegiau-English bilingual—
Norwegian in the left-hand column of the introduction. English in the
right, Norwegian and English glos.ses to the Lappish woids. etc. The
vast majority of Finno-Ugrian philologists will agree with me that the
second language should have been German and not English. In the
/first place German, which has always been the standard ]yeltsprache
of the Finno-Ugrian Society and of Finno-Ugrian philology generally,
is understood much more readily than English by the majority of
Finno-Ugrian philologists, particularly in Hungary, but also to a very
large extent in Finland and Estonia. In the second place there are at
the moment no standard technical terms for Finno-Ugrian philology
in English as there are in German and Norwegian. Thus, on p. lii of
vol. i. the Norwegian term hladievcksling (= German SlufenwecJisel)
is translated by English stage-alteration. This Engli.sh term is
unintelligible out of its contc.xt and has no currency whatsoever —
the usual translations of StafeinrecJisel are the much more readily
intelligible consonant{al) ablaut or consonant{al) alternation.
But these are minor points in no way affecting the merit of a really
great and outstanding work.
Ai.an S. C. Ross.
Chaxdba-kaxta Abhidhax. a comprehensive dictionary of the
Assamese language, pp. xxx -f 1 -f- 1044. Assam Sahitya Sabha,
Jorhat, 1933.
The Assam Sahitya Sabha. of Jorhat, has recently compiled and
published an etymological dictionary of the Assamese language, the
236
REVIEWS OF BOOKS
Chandra-kanta Ahhidlmi}. This is the third dictionary of Assamese
which has appeared ; the first was by Bronson, compiled with the aid of
Jaduram Deka Barua which was published by the Baptist ilission
Press in 1867 ; and the second by the late Hema Chandra Barua,
which was published by the Government of Assam in 1900. Bronson's
dictionary contains words written phonetically and etymologies are
wanting. Notwithstanding it served a useful purpose at the time, and
Bronson, and others of the Baptist llission in Assam, are entitled to
all praise for having provided what was then the only good vocabulary
of the language. The Hema Kosha, on the other hand, is an
etymological dictionary and, though not so full and not containing
so many references and quotations from Assamese puthis as the work
now before us, showed a great step in advance and represents what
must have been almost a life-long effort on the part of Srijut Hema
Chandra Barua, who was a finished Assamese, Sanskrit, and Prakrit,
scholar. The learned author unfortunately did not live to see the
publication of his great work, which fell to the lot of others. The
writer remembers the late Srijut Dulal Chandra Chowdhary bringing
him the manuscript of the dictionary and asking whether the Assam
Administration would publish it, a wish which was gratified, for thanks
to the late Sir Henry Cotton, the then Chief Commissioner of Assam,
the Assam Administration undertook the publication, the late Hema
Chandra Goswami and the writer being the joint editors of the work,
which duly appeared from the Baptist Jlission Press in Calcutta in
1900. The preface of the Chandra Kanta Ahhidhan states that this
dictionary contains 36,819 words, whereas the number in the Henia
Kosha is 22,386 only. In the preface we are told how the present work
originated, which was in the following manner. Rai Bahadur
Eadha Kanta Handiqui, a retired ci-vil servant of the Assam Govern-
ment, lost two most promising sons, Chandra Kanta and Indra Kanta,
in the most tragic manner, in the course of a few days only, of typhoid
fever ; the elder Chandra Kanta, a young graduate of the Calcutta
University, and the younger, Indra Kanta, who was reading for the
B.A. at Gauhati. It was whilst nursing Indra Kanta from typhoid
that Chandra Kanta himself wus attacked, the result being that both
boys fell \dctims to the dread disease. The distracted parents, after
an interval, decided to erect a hall at Jorhat as a memorial to their
two sons, and, further, to perpetuate their memory by arranging for
the compilation and publication of an As.samese dictionary, also the
preparation of a Buranji or history of Assam. The hall at Jorhat,
CHANDRA-KANTA ABHIDHAX
237
called the “ Chandra Kanta Handiqui Hall ”, was opened by Sir John
Kerr, the then Governor of Assam, on the 2nd December. 1926. The
second memorial is the Chandra-kanta Ahhidhan. the subject of this
notice. Sir John Kerr in his opening speech said : ” Xo more suitable
method could have been found by any Assamese father for honouring
the memory of his Assamese sons ; and I have no doubt willing and
able workers will always be forthcoming to assist in carrying out the
benefactor’s wish, and that the Institute should become a great centre
of Assamese culture.” The Chandra Kanta Handiqui Hall at Jorhat
is now the headquarters of the Assam Sahitya Sabha, the Society
which has done so much in preserving the purity of the Assamese
language and which has been responsible for the compilation and
publication of the great work now before us. Rai Bahadur Radha
Kanta Handiqui by making the large grant of Rs. 30,000, for com-
memorating his two sons, has conferred an inestimable benefit on his
fellow countrymen by giving them a %'eritable treasure house in the
shape of the dictionary, and all concerned in its production are heartily
to be congratulated on its appearance. To the writer of this notice
the pleasure of studying the many pages of this volume has been great,
more especially the etymologies and the numerous extracts from
Assamese puthis it contains. Further, the excellent preface by
Srijut Deveswar Chaliha and the learned introduction by Srijut
Devananda Bharali, give additional interest and value to the work.
In this brief notice it is not intended to offer any criticisms or review
of the work as a whole. This must be left to competent scholars ;
I refer in particular to the introduction by Devanda Bharali. This
notice is merely one of appreciation by one who was long connected
with Assam, who during his period of ser^'ice there did all within his
power in the course of his official duties, to study the Assamese langua :e
and literature. That the work of compilation and publication is truly
an indigenous effort on the part of the Sahitya Sabha. thanks to the
generosity of Rai Bahadur Radha Kanta Handiqui. is an indication
of the cultural advance of the Assamese people, and is an additional
subject for congratulation.
The dictionary, which is well printed and bound, is priced at
Rs. 7, and is obtainable from the Assam Sahitya Sabha at Jorhat.
P. R. Gurdox.
238
REVIEWS OF BOOKS
MaXICHAISCHE HaXDSCHRIFTEX DER hSxAATLICHEX MuSEEX BeRLIX.
Herausgeffeben ini Auftrage der preussischen Akademie der
Wissenschafteii unter Leitung von Professor Carl ScHinoT.
Band i. Kephalaia. Lieferung 1 2. pp. 50. Stuttgart : Yerlag
W. Kohlhamnier. 1935. RM. 12.
The Prussian Academy of Sciences has undertaken the publication
of the Manichean paptui. To this good fortune we owe the early
edition of this first volume. The form is that of the volume of
3Ianichaische Homilien of the collection of Chester Beatty, published
last year. Here also the work is by H. J. Polotsky and of the same high
standard. Many pages are fragmentary, but many also are happily
complete or almost so. In this volume are contained chapters i-x\u
of the Kephalaia, of which chapters iv and vii had been earlier
published and interpreted in SBA If’., 1933, Ein Mani-Fund in Agi-pten.
M’e have here from the unprejudiced Manichean standpoint inter-
pretations of the manifold aspects of the Manichean system, so of
the parable of the good and bad trees, the four fishers, peace, the right
hand, the salutation, the fourteen aeons of Sethel, and many other
topics in chapters of various length. The allusion on p. 12 to Zarades
(Zoroaster) places him after Bouddas (the Buddha) in time. On p. 7
there is a reference to Zoroa-strians reading their books, but this occurs
in a schematic e.xplanation of the fallibility of earlier religions, which
is attributed to the fact that not the Teacher but the disciple recorded
the Teachings. The allusion therefore loses .some of its importance,
The great passage of this volume is that on the mission of Manichaios
himself, pp. 14 ff., which was earlier quoted in translation in iSBM IF.,
1933 (cited above). It agrees with the Arabic account of al-BairunL
The technical production of the book is e.vcellent. The Coptic
is in the clear new ti-pe. Footnotes are few but judiciouslv support
the interpretations. H. W. B.
Svllabaire Hittite CrxErFOR.\iE. Par Lons Delaporte. Libraire
d'Arnerique et d'Orient Adrien Maisonneuve. Paris. 1929. Frs. 30.
This book, forming part of 31 . Delaporte's Manuel de Langur
Hiflitr. intended for students entering upon this study, contains a short
preface on the cuneiform tiqie of writing and the lists of signs, in two
arrangements, according to the form of the sign and according to its
syllabic value. A table of variants of the elements forming part of the
cuneiform signs is appended. It is a useful and indeed necessary part
of such a manual. jp g
CAUCASICA
239
Caucasica. Fasciculus 11. Herausgegeben von G. Deeters. Leipzig ;
Yerlag Asia Major G.M.B.H., 1934.
Four monographs are comprised in this fasciculus. One is concerned
with Iranian. Ossetische Studien, by K. Bouda. They contain a
useful contribution to the definition of the meanings and history of
words and to Ossetic sjuitax. The author has treated komdaryn.
komjTi, cast, siir, umui, the deictic use of pronouns, kanyn in
intransitive phrases, Ossetic-Caucasian relations, old material in
Ossetic, the Accusative, astau, the postposited adjective, and group
inflexion. Under the Ossetic-Caucasian relations it is not altogether
clear what connections are intended. In the case, for example, of
Oss. Mf ‘‘ fish ", it is necessary to remember Sogdian kp {— leaf) and
Saka kara- with the same meaning. On the Cerkes there are two studies.
Prmce N. Trubetzkoy has recorded his recollections of two visits to
the Cerkes, so far as they have remained in his memory after loss of
his notebooks. He has here given chiefly translations of tales written
down during his %'isits. The second article is by G. Deeters, a linguistic
analysis of a Cerkes children's learning book. E. Bleichsteiner in
“ Die Kaukasischen Sprachproben in Evliya has examined the words
and phrases recorded by Evliya (jVlebi, and has succeeded in most
cases in tracing their equivalents in modern Georgian, Mingrelian,
Abyaz, and Cerkes. The words are important evidence to the three
non-literary languages three hundred years ago.
H. AY. B.
ETrDE.8 COMPARATIVES SUR LES LANGUES CAUCASIEXNES DU NORD-
ouEST (Morphologie). Geoeges Dumezil. Librairie d’Amerique
et d 'Orient Adrien-Maisonneuve. 1932. Frs. 80.
Textes populaires Ingus. Recueillis par M. Jabagi. Traduits
commentes et precedes d’une introduction grammaticale par
G. Dumezil. Adrien-Maisonneuve. 1935. Frs. 30.
The comparative studies of the first of these two books are confined
to the morphology with occasional indication of comparative
phonologv. Morphology before a basic phonology has been worked out
has not been approved or possible in Indo-European comparative
grammar ; it is exposed to the risk of accidental similarities. It is
therefore doubtful if arndhing more than similarity of general structure
or of s}'ntactical use can be shown in this way.
The eighteen Ingus texts are welcome. Jahagi had collected
traditions, songs, and proverbs of the Ingu.s, both from an interest
240
REVIEWS OF BOOKS
in folklore and as a service to bis own people. They are thus popular
texts. The editor has given the texts with interlinear translation
accompanied bv a freer tran.slation at the end.
H. W. B.
Die Traxier. Arthur Christexsex. Handbuch der Altertums-
wissenschaft. S'* Abteilung, 1'"' Teil, 3'*'' Band. 1933.
Professor A. Christensen has in one hundred pages written a most
excellent account of the origins and of the culture of the Iranians to
the end of the Achaemenid period, with an Appendix on the period
between Alexander and the end of the Sasanian Empire, Useful
bibliographies are given throughout. He has treated of the early
period, the reform of Zoroaster, the Iranians of the west and of the
north. The sketch of the oldest and least known times and of the
activity and beliefs of Zoroaster probably approaches as near to the truth
as present information permits. It is admittedly incomplete evidence
For the Western Iranians more is to be expected when the Elamite
tablets, discovered at Persepolis, are published. The recent discovery
of references to Cvrus I in Babylonian records encourages hopes of
further information from that source. The author has carefully weighed
all the available information, which, as he notes, is largely preserved
in non-Iranian books. Happily Greek authors supply the want created
by the lack of indigenous literary sources. Eight excellent plates are
included.
H. W. B.
A j-IBETAx-Exglish Dictioxary. With special reference to the
prevailing dialects. To which is added an English-Tibetan
Vocabulary. By H. A. J.Vschke. London: Kegan Paul, Trench,
Triibner and Co., Ltd. Reprinted 1934. 42s.
^lO Tibetan dictionary is exhaustive. Such a one is greatly needed.
Meantime it is good to hai'e available once more at a reasonable price
Jaeschke s very useful dictionary. It had become scarce and expensive.
The contents are unchanged, only the format has been reduced.
Daily use has demonstrated its handiness. With the help of this
dictionary and that of S. 0. Das it is possible to work at Tibetan
texts, but the omissions of both are often disturbing. Reading recentlv
the Siddhasara of Ravigupta in the Tanjur, I have met often words
or phrases which neither elucidated. Thus li-zir-ha, corresponding to
sat! - zedoary " of the Sanskrit text, is in neither. Skf. rajanl-dvaya
“ curcuma longa and aromatica ” is rendered by sgyer-ba dan skyer
EEVUE DES ETUDES ISLAMIQUES
241
Ichanda dan, of which sicyer hhanda, omitted by Jaeschke, is explained
by S. C. Das as a confection of skyer-pa. The word for “ coriander ”
is hu-su in the dictionaries, the medical text uses ho-su. Neither
dictionary has kha-lun. which translates Skt. matuluhga " citrus
medica ” (it occurs in the Mahavvutpatti), nor sin he-ron used for
Skt. trivrt turpethum The word ts‘on-mo-ste)t, omitted in S. C.
Das, and stated in Jaeschke to be an unknown metal, translates
Skt. trapvs “ tin " (also in the Mhvy.). Skt. aina- deer's (flesh) " is
rendered by khar(a) go sol gyi sa. Further sin-tu zi-zi-por gyar-pa
translates drster dkulatd bhrsam. Future compilers of Tibetan
dictionaries may also expect to find much of use in the Chinese
polyglot dictionaries.
H. \V. Bailey.
Revue des Etudes Islamiques. Publiee sous la direction de
L. Massignox. Tome VIII (Annee 1934). 4 cahiers. pp. 512.
Paris : Geuthner. Subscription price 100 francs per annum.
This latest volume of the R.E.I. is remarkable not only for the
variety of its documentation on Islamic subjects, ranging from Barbary
to Japan, but still more for the large proportion of articles dealing
with new or almost untouched problems. In this respect it holds up
a mirror to the present state of Islamic Orientalism, with results at
once encouraging and disheartening. Such preliminary articles as
those of Albert de Boucheman on the rivalry between two semi-
sedentary nomad groups in SjTia (pp. 9-58). of Jean Sauvaget on the
stages of the growth of Damascus as a city (pp. 421-480), and of .A. N.
Poliak on the popular risings in Egypt in the Mamluk period (pp. 251-
273) all represent pioneer work in important fields of historical study.
While they hold out the promise of valuable extensions of knowledge
in the future, one cannot help feeling a touch of despondency at the
little way we have gone and the long way there is still to go. Somewhat
similar reflections are evoked in another field by Professor Massignon’s
interrogative memoranda on the formation of abstract nouns in Arabic.
The note of pessimism is openly sounded by Tadeusz Lewicki in
publishing some new texts in Old Berber from an anonymous twelfth-
centur}" work (pp. 275-306) and by E. Borrel in regard to the popular
poetry and music employed in the former secret religious assemblies
of certain Turkish tribes (pp. 241-250). the first because the material is
so scantv. the second because the study of this subject, so important
in its sociological bearings, has begun ten years too late.
VOL VIII PART 1
16
242
REVIEWS OF BOOKS
This bent towards social investigation, however, w’hich is
responsible for much of its element of originality, constitutes, together
with summaries of current acthuties, the special feature of the B.E.l.
and gives it a place apart amongst Islamic periodicals. Most of the
remaining contributions also fall within one or the other category.
A summary of two articles by Gordlevsky on the artisan corporations in
Turkey (pp. 79-88), Paul Marty's article on the recent difficulties of the
silk-weavers in Tunisia (pp. 223-240), and 3Ime. Jeanne Jouin's
illustrated survey of the traditional dress of the women of SjTia and
Palestine (pp. 481-505) continue the tradition of sociological study, while
Claude Cahen contributes a corpus of useful geographical and historical
material on medieval Mesopotamia from the unpublished work of Ibn
Shaddad (pp. 109-128), and Lewicki an analysis of a fifteenth-century
Tbadi chronicle, the Kitab as-Siyar of ash-Shammakhi (pp. 57-78).
Current developments are represented by a variety of summaries ;
of the Firdausi celebrations in Iran, by Professor Masse (393-420),
of extracts from the Ru.ssian press relating to Islam in Central Asia and
the Far East, by J. Castagne (pp. 89-108), and of the reports made at
the Congress of Writers of the U.S.S.R. on the literature of the Central
Asian republics, by B. Nikitine (pp. 307-381). A brief account of the
Muslim community in Finland (pp. 1-8) and an obituary of Ahmad Zaki
Pasha (pp. 388-392) are contributed by Bichr Fares, and the abridged
translation of 'All ‘Abdar-Raziq’s Isldm and the Bases of Government,
begun in the previous volume, is concluded (pp. 1G3-222). And. by no
means the least important feature of the volume, the series of Abstracta
Islamica is continued in a long and valuable section, compiled bv
Dr. Paul Kraus, of works relating to Islamic medicine and .science
(pp. 126-162). H. A. R. G.
The Holy Qur-Ix. English Text and Commentary (with Arabic
Text). By A. Tusi f .\lI. Part I containing the First Sipara.
Imperial 8vo, pp. xx -f 56. Lahore, 1934. Rs. 1,
L.a. Sages.se Cor.\xique Eclairee par des Merset.s Choi.sis
Par IMahmoud Mohtar-Katirjoglou. 8vo, pp. 261. Paris ■
Geuthner, 1935. 22 frs.
The late General Mahmoud Mohtar Pasha, in his introduction to
the second of these books, remarks that ■■ a sacred text, if it is to live
afresh in another tongue, requires to be given life hv pens which are
imbued by that same life.'’ The be.st translation will be one made by
a believer in the sacred and unique character of the text, and will not
THE HOLY QTTRAN
243
always be the most literal translation. This is a liberty of which
Mohtar Pasha himself has taken full advantage, his object being simply
to unite in one short volume the essential elements of the Qur'an "
by the omission of verses and parts of verses amounting to about
four-fifths of the whole.
The translator who sets out to reproduce the entire text, on the
other hand, cannot afford to wander far from the literal rendering.
For in a sacred text the minutiae may. ex hypothe.si. be of the mo.st far-
reaching importance. Mr. Yusuf ‘Ali, too, of course, has aimed at
something more than a bare rendering. In addition lie has tried,
greatly daring," to reproduce something of the rhjdhm of the original
by the use of unrhymed free verse, each line of which begins with a
capital. The rh}i;hmical quality is certainly present, but the reviewer
must confess that the effect is, in his opinion, altogether unhappy.
The translation itself contains many happy turns of phrase, but is
marred by several loose and sometimes fanciful renderings.
In both works, however, the actual translation is perhaps of less
importance than the commentary. \\’e are in reality dealing with two
highly individual recensions (the word is not too strong). Each has
the interest of presenting, in Mr. Yusuf ‘Ali's words. the picture
which their own mental and spiritual vision presents to themselves."
and each is the work of a man of wide culture, with long training and
experience, not in academic or theological seclusion, but in the busy
life of politics and public serxTce. Both are modernist, in the sense that
thev reject the formalism of the rigid orthodox tradition, and maintain
the characteristic modernist doctrines, such as free will and the
symbolical interpretation of legendary narratives and of the
descriptions of the future life. There is. however, a significant difference
between them : Mr. AMsuf "Ali's commentary is mainly '' practical "
and homiletic, whereas Mohtar Pasha shows rather a sufistic-
philosophical tendency. Both are well worthy of the attention of tho.se
interested in the modern development of Muslim thought.
H. A. R. G.
Histoire des Ceoisades et DU Royaume Frax'c de Jerusalem.
II ; Monarchic Franque et Monarchic Musulmane ; L'Equilibre.
Par Rexe Grousset. 8vo, pp. iv -f- 921. Paris : Plon. 193.5.
100 fcs.
M. Grousset's second volume, which, notwithstanding its bulk,
follows the first at an interval of only a few months, deserves more
244
REVIEWS OF BOOKS
than a passing notice, ^\^lile it follows in the main the lines of the
previous volume (reviewed in Vol. VII, pp. 981-3). it shows also some
significant differences in detail. As the events of this period are less
dispersed and lend themselves more easily to unified treatment than
in the earlier period, there is less ov'erlapping and a corresponding
gain in clarity. In the second place William of T\Te now becomes the
primary Western authority, and though 31. Grousset continues to
prefer the old French version, he calls attention to discrepancies
between the original and the translation where they are significant.
This is all to the good, and the numerous quotations take on in con-
sequence the character of first-hand or eye-^vitne.ss accounts, which in
the first volume they did not always possess. There is also a marked
change, not of outlook but of stress, as the author gets more fully into
his stride ; his personal judgments are more intrusive and more heavily
underlined. Partly, no doubt, becau.se in his defence of the later
Capetians " he is aware that he is arguing a case ; no historian has
doubted the abilities and qualities of the first two Baldwins, but
Fulk, Baldwin III, and Baldwin IV require the support of a vigorous
advocacy, which involves him even in some polemics against
Chalandon. These, and various other matters in which he expresses
a personal view, belong to the prordnee of 3Vestern medievalists,
and cannot be discussed here. But this tendency to emphasis has
resulted in some lack of balance, both in the evaluation of the
moral factor (of this more below) and in the treatment of such
episodes as the battle of 3Iontgisard (“La plus belle victoire des
Croisades," 654 sqq.).
The leading ideas in this volume are firstly that the monarchy
was the sheet-anchor of the Latin states — the case for which is argued
passionately and on the whole persuasively ; and, secondlv. that the
Frankish occupation was fundamentally an e.s.say in colonial expansion,
to which the “Crusading ideology" was in direct antithesis (e.g.
pp. 247 and 615-17). The weight laid upon these factors will be best
appreciated from two quotations : —
La solution [i.e. of the struggle between the Turks and the Franks]
allait presque entierement dependre de la valeur respective des
institutions monarchic|ues chez Ics deux adversaires ” (p. 366).
■■ La, oil il eut fallu une veritable Volkerwanderung. une
immigration rurale massive, il n'y eut plus que des vovages de paladins
et de marchands— chev-aliers-moines des Kraks. armateurs et banquiers
italiens des Lchelles libanaises, bref une colonie sans base
HISTOIEE DES CROISADES
245
clemograpilique et rurale, le Pondichery de Dupleix au lieu du Canada
de Montcalm, la colonie qu’un desastre militaire aneantit au lieu de
celle qui survit meme a la conquete etrangere " (p. 865).
These ideas largely dominate M. Grousset's exposition : he cannot,
for example, forgive the B 3 ’zantine empire for its hesitant and finalh'
hostile attitude which allowed the destruction in Anatolia of the
Crusades of 1147 with their multitudes of possible future colonists
(Appendix IV). Often, too, they lead him into speculations in which
the broad overriding facts (although M. Grousset is by no means
unaware of them) are apt to be lost to view while he expatiates on the
moral of this or that episode in the light of the doctrine of monarchical
institutions or of “ le fait colonial A characteristic example will be
found in his impatience at the missed opportunity of erecting a
'■ Frankish Countv of Middle-Orontes ” in 1157 which, he maintains,
if given to Thierry of Flanders, would have secured the support of the
Flemish cities (p. 384).
The presentation of the oriental aspects reflects, as in the previous
volume, the absence of preliminary orientalist researches. There are,
it is true, man v passages of shrewd analysis and of just appreciation, but
the oriental figures and background are not solidlv built up ; even in
dealing with Saladin (for whom the fullest documentation has hitherto
been made available to the Western student) the reader is not prepared
for the final catastrophe, which is, indeed, a.scribed almost entirely
to the negative factors of weakness in the kingdom. In so far as the
utilization of the available oriental sources is concerned, however,
these is little indeed of importance that has been overlooked, ^ but in
spite of the author's caution Ibn al-Athlr has again proved something
of a snare, especially in relation to Eg^-pt. Thus the situation in Egjpt
prior to the siege and capture of Ascalon in August, 1153 (pp. 349-353),
is somewhat misrepresented. The attack on Ascalon was not caused
directl}’ by disorders in Egjpt. Ibn Sallar was, on the whole, a capable
and energetic ruler, who reacted to the Franks' fortification of Gaza
first by a naval raid in the summer of 1151 (related in the Dam.
ChroH., 307-8. but overlooked by M. Grousset) and secondly by
1 The followina point.s are perhaps worth mentioning : pp. 70-1 ; Kamiil ad-Din's
statement cannot .«.tand again.st the (oneordant testimony of lim al-Qa!anisf, William
of Tvre, and Ihn al-Atlur : p. 367 ; the truce between Xur ad-Din and the Franl-s
was tirst made in llo.j (Vum. Chron. 32i) ; pp. 416-17 : according to the Dam. Chron
(3.53) JIanuel was negotiating with Xur ad-DIn even before his advance, and this
seems to be contirmed by the presence of Xur ad-Din'.s envoys at Mamistra, as
mentioned on p. 403.
246
KEVIEWS OF BOOKS
endeavouring to form an alliance with Xur ad -Din. ^ The siege of
Ascalon was formed in January, 1153, whereas Ibn Sallar was
assa.ssinated only in April. This may have contributed to the conquest,
but even the new government in Eg^-pt strongly reinforced the besieged
garrison. On the other hand, in preserving what may be called the
conventional view of the events in Egj’pt leading up to and following
the establishment of Saladin, again based mainly upon Ibn al-Athlr.
M. Grousset follows probably the .safest plan, in the absence of a
detailed study.
Mention of Saladin brings us to the crux of the problems associated
with the Muslim counterstroke and the fall of Jerusalem. M. Grousset
as has been seen, finds the explanation firstly in the recovery of the
monarchical institution on the Islamic side, coinciding with its decline
on the Frankish side, and secondly in the failure to place the Frankish
colonization of Syria on a sound numerical and agricultural basis.
It may be questioned whether the second was an effective cause in any
degree. If there is any virtue at all in historical reasoning, nothing
can be more certain than that, however strongly Syria might have
been colonized, the enormous reserves of strength possessed by the
Muslim East — decapng as it was — were more than sufficient to have
put an end to the experiment sooner or later. How far the very
modern-sounding colonial concepts of M. Grousset can be applied
to the French of the twelfth century is another question, which again
falls within the sphere of the M'estern medievalists.
The first reason is more .substantial, and that it contains an
important element of truth can hardly be doubted. Yet a great deal
of misunderstanding has been cau.sod by the transference of Western
conceptions of monarchical iustitutions into medieval Islamic
society. M estern historians generally .see in the sequence ZankI —
Xur ad-Din- Saladin a .steadily increasing mobilization of military
force based upon a gradual expansion of territorial holdings and
simultaneous increase of .sovereign status. The historv of the crusades
becomes a sum in simple addition and subtraction. M. Grousset
does himself some injustice by allowing it to appear as if he, too,
held this view. For in his close examination of the Oriental sources
he has not failed to notice and to comment upon several of the passages
that hint at, rather than di.sclo.se, the moral revolution which underlay
the political action, and which was the true decisive factor.
1 Itn .Sallar was .a .Sunni and was supported t>y the .Sunni party in Etrypt {.lamiU
.id-lJIn al-Halabi, B.il. Or. 36.S.'), fol. 87a).
HISTOIRE DES CROISADES
247
Professor Massignon has pointed out in another connection that
movements in Islam are generally “ in the nature of sudden flashes,
momentarily violent but transitory There is little of gradual and
organized building-up, but instead there is a long period of secret
preparation during which the social conscience is aroused, and the
movement suddenly comes to maturity in a “ general mobilization
for the struggle which seeks to vindicate by force of arms the neglected
rights of the Divine Law It was precisely with such a movement
that the Crusaders were faced in Syria. There is no space within the
limits of a review to develop this theme at length, but the essential
points may be briefly stated. It is admitted that the fundamental
cause of the weakness of the Muslims was the spirit of particularism
and local jealousy which reigned in S^nia and Mesopotamia, and which
ofiered opportunity of personal aggrandizement to ambitious princes
and governors ; that any political unity created merely by force was
unstable and would end, like the empire of Tutush, in disruption ; and
that the creation of a true political unity had to wait on the growdh of
a psychological or spiritual unity strong enough to master — even if
only for a moment — the obstacles of regionalism and private interest.
It was this that made the path of recovery so slow and painful. Only
thirty years elapsed from the arrival of the Crusaders to the rise of
ZankI ; sixty years had to pass between Zanki's appointment to Mosul
and the battle of Hittin. It ^vas no material cause which delayed this
consummation. Had Zanld and Zanld’s army and Zanki's subjects
been animated by the spirit of Saladin and his men, gallantry and
statesmanship would have been of little avail to Fulk and his successors.
But that spirit was lacking. To Ibn al-Athir and his followers ZankI
may have been the Champion of Islam " ; to his contemporaries —
as L'sania and Ibn al-Qalanisi bear ample witness — he was a " blood
and iron prince, crafty, cruel, ruthless, and as faithles.s as a Frank. -
Yet, had Zanki himself been other than he was, the climax would
still have been dclaved, because the dn'iva — the moral call — had not
vet penetrated the social conscience. In such nmvements the impulse
has generallv to well up from below, establishing itself in certain circles
and thence expanding its range of influence. We can dimly discern
' Whiiher IsJam ? Lonclun (Golluncz), 1932, pp. 77-8.
- M. Grousset admits thi.-. but imagines that the Muslim •norld discovered it only
in 1139 (pii. 129-130). It is implicit in the Dam. Chron. from the outset of
Zanki's eareer, however, and explicitly stated apropos of llS-j {Dam. Chron., 230).
while Usama’s grim stories of Zanki and his generals tell their own tale.
248
REVIEWS OF BOOKS
this process in SjTria, firstly in the attitude of the religious leaders ;
the increasing stress laid on the defence of Islam, the cautious support
of Nur ad-Din, the enthusiasm for Saladin ; how the continued
insecurity of life and property evoked amongst all classes a
psychological unsettlement * which, in the manner of the Orient, took
on a religious expression ; how this religious reviv'al, by setting a new
value on the common faith, brought to an end the crippling opposition
of Sunni and Shidte (except in the mountain regions) — quite apart
from the official " Sunni reaction " manifested in the building of
madrasas, etc. — and formed a bond of union between the Muslims
of inner S^-ria and those who lived under the rule of the Franks. -
But these spiritual aspirations could not issue as a political force
until they found a leader capable of personifying them and guiding
their action. The history of the Counter-Crusade is therefore built up
of the interplay of the moral factor and the changing political situation.
So far from any mechanical progress of the monarchical institution
on the Muslim side, it is instructive to observe how, on the death of
each successive leader, the structure which he has created by his
efforts immediately falls to pieces. The division of Zanki's possessions
actually threw Muslim Syria back to its former condition of political
fragmentation, and Nur ad-Din had practically to build up his own
structure afresh. The death of Nur ad-Dln again jeopardized the
Muslim cause by the imminent dissolution of the Zankid structure,
and once again the “ monarchical institution " had to be rebuilt from
the foundations. So little was there of steady, natural expansion. The
most one can say is that each of the three had the way eased for him
by the achievements of his predecessor : ZankI by Aq-Sunqur al-
Bursuql, Nur ad-Din by ZankI. Saladin by Nur ad-Din. But each had
to bring his own personality to the task of rebuilding, and in each case
the inner spirit was different — not less so between Nur ad-Din and
Saladin than between ZankI and Nur ad-Din. The difference between
ZankI and Nur ad-Din was that the latter, conscious of the growing
moral revolution, endeavoured to utilize it and to supplement political
unity by moral unity (with no little success, as M. Grousset has shown
on p. 460). Why the Counter-Crusade did not come to a climax under
1 Probably strongthened by the series of severe earthquakes which began in 11.56.
- On the bearing of tlie passage.s from fbn Jubair, quoted liy M. Grousset on
pp. 7.52-7, .see now Claude Cahen, in Indigenes et C'roises ”, Syria, 1934, 3.56-360.
Moreover, the Western writer is apt to underrate the effects of the moral degradation
felt by the Muslim under non-JIuslim rule, and .so vividly portrayed by Ibn >Tubair
himself (G.M.S. ed., 307).
HISTOIBE DES CEOISADES
249
his leadership still awaits explanation. It mav have been the still
imperfect moral unification of the people of S\Tia, or the legacy of
Zanki — the character of his army, and the political instability resulting
from the partition of Mosul and Aleppo, at once concealed and main-
tained by Xur ad-Din's loyalty to his kinsmen. Or it may have been
some strain of Zankid hardness and ambition in Nur ad-Din himself ;
for while his real piety and nobility cannot be called in question
(even though all our information comes from the pages of eulogizers),
he was, after all, a Turkish general, born to power, and the continuator
of his father's work.
But Saladin was a child of the dawa itself, moulded in childhood
and youth by its aspirations and ideals. He was not remarkable
either as strategist or statesman ; as scholar and ascetic he began,
and scholar and ascetic he remained all his life. M. Grousset seems to
question whether Saladin was really an outstanding figure ; he is
inclined to put him on a level with the Zankids and other atabegs, and
he accepts Ibn al-Athir's partisan insinuations ^ and assertions.’^ A"et
the one charge brought against him — his apparent disloyalty to the
house of Nur ad-Din and his determination to make his supremacy
acknowledged in North SjTia and Mesopotamia — was the fruit of a
just appreciation of the situation. For the weakness and division of
the Zankid house threatened to plunge S}Tia again into disunity and
discord and to betray the hopes of Islam. The Zankids themselves
pointed the contrast ; their reliance upon Frankish support
(pp. 622 sqq.), a commonplace of Syrian politics in 1120, and already a
“ regrettable necessity ” in 1150. was a thing not to be tolerated in 1180.
The lesser personal loyalties which opposed Saladin in the beginning
had in the end to give way, and were merged into the general movement
as the aspirations of the Muslims of Syria found in him their ideal
representative. Besides the story of the garrison at Harim (p. 720),
other instances could be quoted as widely separated as Ibn Jubair's
account of the prayers at Mecca (G.M.S. ed., 97) and the change of
1 See. c.g.. ])p. o91-2. But how could Ibn al-.\thlr know what Xajm ad-Din had
said to his son in a private conversation ? The whole story is typical of Ibn al-Athir’s
imaginative excursions.
- Amongst these the statement that Saladin’s expeditions into Nubia and Yaman
were intended to prepare a pos.-ible place of retreat (p. 593) has been widely accepted
by historians. Yet it might very plausibly be argued that their purpose was firstly
political, Yaman being the headquarters of the Fatimid propaganda and Nubia a
centre of Fatimid resistance (cf. p. 597), and secondly economic, for the protection
and control of the trade-route via '.Aidhab and the Red .Sea. the chief artery of Egypt 's
commerce.
250
REVIEWS OF BOOKS
tone in the chronicle of Ibn al-AthIr himself. M. Grousset has scarcely
realized the full implications of this process, but he sees clearly the
signs of the coming outbreak in the reaction to Renaud de ChatUlon’s
exploits in the Red Sea (pp. 734-5). The last consideration which
might oppose itself to the general will, the economic interests of the
SjTian merchants, was swept away by Renaud ’s perfidy (pp. 776-8) ;
barely six months later the kingdom of Jerusalem was in dissolution.
Saladin's achievements are not to be explained by any theory of
monarchical institutions. The outcome of the campaign of Hittin
and of the Third Crusade was determined by morale, zeal, and
enthusiasm for an ideal. The Franks of SjTia had indeed sad need
of some infusion of the early ‘‘ Crusading ideology '. For the Crusades
were in essence a contest of faith, not of arms.’^ AU the material
conquests of the First Crusade were the result of the first uprush of
zeal and devotion ; when that gave way to “ le fait colonial ”, it was
by an equal uprush of zeal and devotion on the other side that they
were taken away.
H. A. R. Gibb.
Book of Trea.sures. By Job of Ede.ssa. Edited and translated by
A. Mixgaxa. pp. xlviii + 470. Cambridge : W. Heifer and
Sons, 1935. 42.'?.
It is an exaggeration to call this book an encyclopaedia of
philosophical and natural sciences as taught in Baghdad about
A.D. 817 ”. as is done on the title page. It sets out to .show that the
world is formed of four elements with examples, taken from physics,
zoology, and psychology, of the effects of the various combinations of
tiiem. Teleological rea.sons for mauv of the phenomena described
are also given and the book end.s with an account of the ne.xt life which,
though also formed from the four elements, is quite different from the
present world. Tlie author taught that hell was not punishment by
lire but absence from God.
The facts — real or a.ssuincd -are common to the Muslim world ;
^ i>y faith I tio not mean fanatiri^^m ; nothing is more charaoteristie of this conte.-^t
than the mutual courtesy and re>peft displayed by the combatants. That things
might have been very different i.s >hown by the episodes in which Templars and
Hospitallers were concerned. Fanaticism called out fanaticism in reply, even in a
Saladin, but his generally chivalrous conduct is probably the greatest tribute of all
to the character of the Franks of Syria.
HISTOIEE DES CKOISADES
251
e.g. the three colours of the rainbow, the lack of a backbone in birds,
and the growth of hair from excess matter in the body. The editor
quotes a few parallels from the Firdaws al-Hikmat ; he might have
given many more. The arguments are novel and, though nonsense,
are often ingenious and amusing. The following sample of reasoning
may be quoted : —
“ There are five elements in the composition of man and of the
animals, the fifth element being tbe body, which is composed of them.
When the middle part of the body was being completed, its matter
was pulled upwards and downwards, and reached the extremities
and gave rise to the division of hands and feet ; there each one of the
elements struggled to separate itself from the others, and in this way
the five-fold division of the five powers — that is to say of the four
elements and of the compound that came out of them — took place.
The body, which is composed of a quantity of matter greater than that
possessed by any other single element, perfected the thumb, while
the four other elements perfected the four fingers and the four toes.”
(p. 46.)
The author knew the experience called “ seeing stars”, it seems
that he was influenced by Islam when he spoke of angels carrying the
throne of God, and he attacks some new philosophers, probably the
Mu'tazila, by aSirming that sensations are not essences. The text at
the end of the book is a facsimile of a modern copy of an old
manuscript ; this is always unsatisfactory. The facsimile is on the
whole legible. The editor has to confess that some passages are corrupt
and elsewhere has made many corrections in the notes. Such corrections
are not usually called an apparatus critims. The translation assumes
many other emendations. On p. 92 is a bad example of the method
of translation. A note says that the text is corrupt, but gives no idea
of the e.xtent of the corruption. Tlie translation contains words which
are not in the original and also a beautiful c.xample of homoioteleuton.
In the note on p. xx the word ” reconstructed " should be “ quoted ".
The chief value of this book lies in the light it throws on the history
of science ; therefore one would e.xpect precise indication of all points
on which tin- author differs from Aristotle. The present edition does
not give this.
A. S. T.
252
REVIEWS OF BOOKS
Das Buck dee Alauxe und Salze. Ein Grundwerk der spat-
lateinischen Alchemie. Herausgegeben, iibersetzt und erlautert
von Junius Euska. pp. 127. Berlin : Yerlag Chemie, 1935.
En poursuivant ses recherches sur les ouvrages alchimiques
attribues a Muhammad b. Zakarivya al-Razi (Rhazes) dans les
litteratures arabe et latinej M. Ruska a reussi a elucider un
chapitre interessant de I'histoire de I'alchimie occidentale. L'elabora-
tion de cette decouverte fait Tobjet de la presente etude.
Vincent de Beauvais et Roger Bacon citent a plusieurs reprises
un livre intitule De Alummibus et Salibm et souvent attribue au
celebre medecin Rhazes. II y a quelques ans. M. Robert Steele avait
retrouve et edite le te.vte latin utilise par ces auteurs. M. Ruska
etablit d'abord qu'une .seconde recension du meme texte a ete imprimee
en 1560 par Joannes Garlandius sous le titre De Mineralibus liber.
Les differences apparentes qui se manifestent entre les deu.v recensions
s’expliquent par le fait que ce sont la deux traductions differentes
d’un seul et meme texte arabe. Une heureuse chance a permis a
M. Euska de retrouver une grande partie de Toriginal arabe dans un
manuscrit de Berlin. L edition de ce texte ainsi que de la version de
Garlandius, accompagnee d une traduction en allemand et d’un com-
mentaire, forment la partie principale du travail (p. 39 suiv.). Celle-ci
est precedee d’une analyse minutieuse de la technique de traduction
pratiquee par les auteurs des deux versions latines (dont I'un est
probablement le celebre Gerard de Cremone) et d’un resume du contenu
scientifique du traite. 0 est dans ces chapitres ainsi que dans le com-
mentaire philologique cjui accompagne la traduction, qu’excelle la
methode de M. Ruska appliquee deja a tant d’autres ouvrages du
meme genre litteraire (comp, notamment ses etudes magistrates sur
la Tabula Suuiragdina et sur la Turba Philosophorum). Bienque
1 attribution du traite a Razi soit impossible de nombreux indices
‘ Cf. Uebersetzung und Bearbeitungen von Razl’s Buch der Geheimnisse,”
ap. Quellen und Siudien zur Geschkhte der y aturwksen.schaften und der Medizin,
iv. Heft 3. 1935, pp. 1-87 ; 'Die Alchemie ar-Razis,” ap. Der Islam, xxii (1935),
4, pp. 281-319.
2 Ce qui rend deja suspecte I'authenticite du traite, c'e.st qu’on y trouve cite a
plusieurs reprises le kitdh nl mujarraddt de .Jabir (les versions latinos rondent ce titre
jtuThher denudalorum ou liber de spnlintionihur. respectivement). Rcmarquons d'ailleurs
qu un auteur posterieur pretend que Razi .se serait approjirie le contenu de ce livre.
On lit en effet dans Ic kildb mifatih al-rahma de Tugra’i (MS. Paris, ar. 2614, fol.
109'} la note suivantc : oby.:Jl ^ ^.Achll ..^Ul j JL: 4 jj| Jl;_,
J ';4 ^^4 V Aj} yfj ^1 (sk.) ^1,
AL FARABi DANS l’eCOLE PHILOSOPHIQUE MUSUL.MANE 253
prouvant qu’il a ete compose en Espagne au cours du 11^6 siecle,
son interet pour la connaissance de la transmission de la science
arabe a I’Occident n'est pas moins grand. Son contenu est purement
experimental et releve de la tradition alchimique inauguree en Islam
par les ecrits de RazI et de Jabir. Des traits caracteristiques le
distinguent nettement du type d'alchimie allegoriqtie qiie representent
d’une part le Corpus des Alchimistes Grecs et d'autre part l ecole
egyptienne de I’alchimie arabe.
Paul Kraus.
La Place d’al EarAbi dans l’ecole philosophique musulmane.
Par Ibrahim Madkour. large 8vo, pp. ix + 249. Paris : Adrien-
Maisonneuve, 1934. Frs. 50.
As the title of his work indicates, M. Madkour sets out to e.xplain
the system of al-FarabI and its importance in the history of philosophyL
The author writes of a world of thought which is all but dead ; but he
recaptures its spirit and interprets it with such sympathy and clearness
that the reader is transported to the age of the scholastics.
Beginning with a study of al-Farabi's syncretism, which he
illustrates from his Agreement of Plato and Aristotle, he gives a
systematic exposition of his work and influence, more thorough and
more fully documented than any previous study of this influential
thinker. M. Madkour is undoubtedly right in giving his author the
credit for laying the foundation and pro\ddmg many of the bricks of
the Avicennan building ; but he seems to me to labour the point
unduly, and to overstate the world's neglect of one whom Arabic
writers hailed as the mi/allim al-thdni. It is an exaggeration to say
(p. 2) that al-FarabI is hardly' known in Arabic or Latin, and it may
be conjectured that Avicenna’s confession that the Metaphysics was
unintelligible to him imtil he came across a copy' of al-Farabl's intro-
duction thereto is an acknowledgement of a weightier contribution
from that source than that contained in the few pages of the opuscule
know'n as the Aghrad al-Haklm.
M. Madkour's renderings of al-Farabi are happy', and at least one
silent emendation (p. 21) of Dieterici's text commends itself. It may'
be only' fair to acquit al-FarabI of carelessness in attributing the
so-called Theology of Aristotle to the Stagirite himself, but when
M. Madkour confesses that if that work had not been known to Muslims
its theory' of ideas would have been adopted all the same, he hardly
strengthens the case for the defence. Moreover, as I pointed out in
254
REVIEWS OF BOOKS
The Legacy of Islam (p. 252), the work bears the sub-title The
Commentary (tafsTr) of Porphyry, and includes the statement that it
is a revised version made by al-Kindl. 51. 5Iadkour mentions the latter
fact (p. 135 note), but says nothing about the relation of the work to
Porphyry. I find it easier to believe that Neoplatonism was con-
sciously fathered on Aristotle and a sjuicretism elaborated than that
al-Farabi had omitted to read the title of a work which so profoundly
influenced the course of his thought. It is possible that al-Kindi's
lost writings might throw more light on this question. However,
51. Madkour's discussion of the problems at issue between the two views
of God and the Universe is of permanent value, and advances the
study of 5Iuslim philosophy a further stage. His chapter on the theory
of the intellect is an outstanding example of the value of a study made
bv a scholar who ha\ung been educated in the East has studied also
in the West.
51Lndful of the unhappy di-vusions in Christianity which arose out
of the equivocating of substantia and natura and ovoia and fvais,
I am a little uneasy at the equivocating of s and
(“par sa nature." p. 60, and "par nature" and “ de lui-meme ”,
p. 65). especially in citations from Avicenna. Further, one notes
cA “par .sa nature”, and 3 “par nature”
(pp. 78 and 79).
The WTiter has consulted the best authorities on Western
scholasticism, a subject which naturally is only introduced in its
relation to al-Farabl's system. His work is rounded off with an Index
and a critical bibliography.
There can be no douljt that this book will at once take an honoured
place in the library of 5Iuslim philo.sophy and theology.
Alfred Guill.4lume.
SrKH.ANVAR.\N-i-lR.\.\ DAR 'ASR-i H.vziR. Poets and Poetry of 5Iodern
Persia. Vol. I. By 51. Ishaqi e. pp. 7 -4- 455 + 19. Calcutta ;
Oriental Publi.shing House. 2.5.s.
51r. 51uhamma(l Ishaq (Ishaque), lecturer at the Calcutta
Cniv'ersity, presents in his anthology thirty-three modern Persian
poets whose activities .stretch between the end of the last century and
A.D. 1933. and who are all connected with the great national movement
to which the final impulse was given in the vears 1905-8. Each
item in the anthology consists of a biographical introduction and a
PERSIAN LITERATURE
255
selection of poems. The sketches, written in good and direct Persian,
are a result of M. Ishaque's personal investigation, carried out on the
spot, during his visit to Persia in 1930. They will certainly form the
groundwork of the future histories of modern Persian literature. Their
value is enhanced by the accompanying portraits of the poets which
the author patiently collected in Tehran. As an example of the new
and valuable details found in M. Ishaque’s book, suffice it to mention
his personal interview with the aged satirist Nasim-i Shimal (p. 146)
horn he discovered in his retirement, forgotten and pressed by many
e, . The illustrative poems have also been judiciously selected.
Even a few pages of texts enable the reader in many cases to perceive
the personality of the author. This is perhaps the greatest acquisition
of recent literature. With such honourable exceptions as Iraj-mirza
and Bahar, the mastery of expression of our contemporaries may be
inferior to that of their predecessors of the classical period ; the
antiquated tradition very often even hampers the spontaneous growth
of poetry, but there is one good trait about the modern authors. They
no more live out of Space and Time, remaining impervious to the outer
world. This intrusion of new impressions is rather fatal to lyrics and
mysticism. A romantic return towards the past and even downright
political tendencies prevail in recent Persian poetry and yet this is
still a way out of the impasse in which Persian poetry lingered since
the Safawi times.
E. G. Browne was the first to appreciate the importance of the new
turn of Persian poetry (Press and Poetry in Modern Persia, 1914).
A nicely vuitten and appreciative booklet on the most recent Persian
literature was published in Russian by Chaykin in 1928, and now
M. Ishaque's anthology comes in time to set upon a firmer basis our
knowledge of the repercussions which the present eventful times had
on the most sensitive organs of the Iranian community.
It remains to add that the book of which the author is the editor
is very well printed and bound and that the complete work will comprise
two more volumes which will be impatiently awaited.
V. M.
Persian Literature. A bio-bibliographical survey. Section II,
fasc. 1 ; A ; General History ; B : The Prophets and early
Islam. By C. A. Storey, pp. 235. London : Luzac, 1935. 15s.
Professor Storey’s work is destined not for “reading’’ but for
study and reference. It represents a most welcome addition to
256
EEVIEWS OF BOOKS
E. G. Browne's Literary History of Persia, of which (in spite of all its
importance) the bibliography and completeness of references is not
the strongest point. In the first volume of the work (1927)
Professor Storey treated the rather uninspiring branch of Qur'anic
literature. In his vol. ii he applies his thorough methods to the general
histories in the Persian language. The practical importance of this
part needs no comment : instead of long researches in bulk]t' and rare
catalogues we now, thanks to him, possess a complete and systematic
digest of the data on the authors, their works, the existing MSS.,
and editions with all the nece.ssary references. Yol. iii. which is already
in the press, will treat the still more important branch of “ local "
histories in Persia. After its completion we shall have for the extensive
Persian historical literature a manual similar to Babinger's
Geschichtsschreiber der Osmanen, and when the whole work is ready
we shall finally possess a Persian " Brockelmann ".
The work has been so carefully done and such a mass of catalogues
has been utilized that only longer use of the book mav bring to evidence
some occasional lacunae.
ad p. 66. Some extracts from Gardizi's chapter on the Turks have
been re-edited and translated by Marquart in his Das VoVcstum d.
Komanen (1914) and some more translations by the said author lie
in MS. in the library of the Lstituto biblico pontificale in Rome.
ad p. 75, note 1. The Mashhad MS. e.xamined by A. Z. Validi is
indeed the one containing the history of Oljeytii (Uljaytu) ; along
with the Tashkent MS. it is particularly important for the restoration
of difficult personal and geographical names.
ad. p. 76. The French appreciation of Berezine's edition of Rashid
al-din quoted in note 1 is obviously biased. Berezine's rare work is
a signal service to the history of the origins of the Mongol Empire.
It is based on a careful collation of very good MSS. and the com-
mentary, owing to Berezine’s knowledge of Mongolian, is full of valuable
suggestions,! cf. Barthold in Mir Islama, 1912. I. 60.
ad p. 86. The identity of the “ Anonym of Iskandar ” was
disclosed by Barthold in Comptes-rendus de I'Acad. des Sciences de
rU.R.S.S.. 1927, pp. 115-16 : the author is Maulana Mu‘In al-din
XatanzT. see Daulatshah. cd. E. G. Brorvne. p. 371i4. In a further
article. Yeice ob anonrme Iskendera ” in Bull, de I'Acad. des Sciences
de Vr.R.S.S.. 1929. pp. 165-180, Barthold has, moreover, shorpi that
^ It is inessential that in the chapters on Chengiz-Khan BtVezine deliheratelv omits
the passages on other contemporary rulers, see his translation, xiii,
DIALOGUES IX THE EASTERN TURKI DIALECT
257
the Muntalchab al-tawarlMi-i Mu‘irn (Blochet, iv. 2283, Storey, o.c.
additional 115a) is but a subsequent avatar of the " Anonvm "
presented on 22nd Rajab, 817 7th October. 1414, to Shahrukh : its onlv
difference is that all the pas.sages referring to the former dedicatee
Iskandar have been abridged and his title reduced from hadrat-i
Sultan to aunr-zada. As another example of such a change of the
dedication Barthold quotes the Majmcd al-gharaib bv Sultan
Muhammad b. Darvish Muhammad (sixteenth centurv a.d.) and.
as another instance, the Rdhnf al-^udur may be also mentioned in this
connection.
As regards the transcription of the name of the dvnastv
(p. 87) it seems to be not Kurt but Kart (as marked in the Munis
al-ahrdr MS. and as resulting from a rhyme found by A. Z. Validi).
Xiyal (p. 68) is most certainly *YinaI. Instead of Khicurshdh (p. 117)
and Khu'urshld (p. 152) one ought to write KhimrsluVi and Klnrars/ild
(if not simply Khurshdh and Khurshld !) ; Khdwand (p. 92) is a slip
for Khwdnd. Ardaldn is a better form for Ardildn (p. 146). in spite of
the E.I. The stumbling block of our transcriptions are the Turkish
words in Persian. Personally I consider that they should be rendered
as much as possible according to the original Turkish pronunciation,
so Qoyunli (instead of Quyunli. p. 147), oyhli. etc. Griyoryev is better
than Griyor'ev (p. 108).
V. M.
Dialogues in the Ea.stern Turki Dialect on Subjects of Interest
TO Tra\’Eller.s. Collected and edited by Sir E. D. Ross and
Rachel 0. 'Wingate, pp. 48. J. G. Forlong Fund. '\'ol. XI.
1934. 4.U Crl
Some thirty years ago Sir E. D. Ro.s.s had the chance of meeting
in Calcutta a mullah native of Kaslighar. He used this opportunity
to take down a number of colloquial texts, of which he now presents
eiMit in the form of dialogues conducted by a traveller with the
inhabitants of Chinese Turkestan. The text has been revised by
Miss R. 'Wingate, a practical connoisseur of the Turki spoken in those
parts. The edition must certainly have profited by this collaboration,
but a few places reflect the double conception of some sounds : p. 8 4.
tiishsila, and p. 8 7. tiishurup : p. 40 j. dsViq, and p. 40 4.,. isViq. This,
of course, does not in the least affect the practical value of the dialogues.
In the introductory note the editor sums up some grammatical
VOL. Vril. PART 1 . 1 ”
258
REVIEWS OF BOOKS
peculiarities of his texts. The explanation of the forms like kelliki
as keld-Idr-tki (where iki < ikan) is certainly right, and J. Deny
in his amazingly complete Grammaire de la hnigue Turque, Paris,
1921, § 1361, p. 9-58, has collected numerous materials on the use of
this ■■ dubitative ” form. However, the translation of nechd kiinda
kelliki as " how many days have I’ou been a-coming ’’ seems to have
been influenced by the Ottoman use of iken as a gerund (gelir-iken
“a-coming"), whereas in Eastern dialects ikdn is a “dubitative
past ". Therefore the nuance of the phrase quoted is rather " [I
wonder] in how many days you have come ?
The explanation of -siyan “ in the direction of " by -yan “ side "
is also very probable. Perhaps the form bi{r)siyan < bir-i-si-yan
accounts for the generalization of the form -siyan. Bir-i-si is formed
with a double suffix 3 p.s., but the colloquial elision of the first
affix -i- may have gradually created the impression that the abnormal
-si (after a consonant) is to be connected with the following -yan.
The form no longer felt in its grammatical connection was then
standardized as a sort of invariable postposition.
The Turko-Chinese poem (p. xi) still presents some difficulties.
Verse 2 : ol ay kiiyiga barlp qlshqaray, etc., is translated : “ In this
month people come [?] into the villages [?] crying out : ‘ Carve the
sheeps' flesh ! ’ ’’ As verse 1 announces the coming of the spring
we must take it perhaps that “ that other (winter) month had to go
out {kiiyiga " to the street ”) shouting (in despair 1), etc. Verse 7
exhorts the beloved : “ Come to know me if you desire happiness
{‘dfiyat) . . .” and verse 8 suggests that there is no fear of it being
reported. The Cliinese phrase “ there are no camels here is entirely
analogous to Persian “have you seen the camel ? “ to
which the reply is expected in the negative.
p. 8. Korin originally means “ the .skin of a newly born lamb " ;
p. 13, Khishtln-snrdy “the Tilers' saray", read “sun-dried brick
■sarily": p. 10. note 2, bossipimn. read gossipium \ p. 17, Puslit-i
ptydz ■■ onion-skin can be only a confusion of “skin” and
■ back ; pp. 4 and 1.5 (chay) waza can hardly be ;
more probably it is “ [tea and] delicacies ” ; pp, 34 and 45,
gmlip must surely be qirilip-, p. 20, Jyyl “railway" is Russian
foyezd “ train , and p. 42, chaynek Ru.s.sian cJiaynik “ teapot ”.
V. M.
PROSATEUBS TUSCS CONTEMPORAINS
259
Prosateues TP'ECs coNTEiiPORAixs. Extraits par E. Saussey.
Etudes Orientales publifes par I'lnstitut Fran^ais d'archeologie
do Stamboul, Vol. I. pp. 385. Paris : E. de Boccard, 1935.
M. Edmond Saussey, member of the French Institute at Stambul,
has had the most happy idea of acquainting the public at large with
the present day Turkish literature. The selection, excellently made
and translated, gives an idea of the whole gamut of one of the very
interesting modern literatures which almost entirely breaks with the
past not only as regards the themes, but also the modes of expression.
The Turks have learnt enormously from the French, but their realism
fdls the schemes with purely Turkish material and there is no doubt
that their genuine and original contribution fully merits the attention
of those who are interested in literature in general and those who
want to know what modern Turkey thinks and feels. ‘ ’ Cette litterature
ne chante pas les plaisirs des grands, mais les emotions des masses.’’
says rightly M. Saussey. This literature until now has been studied
chiefly by the Germans (P. Horn. Hartmann, Hacht. Duda) and
Eussians (Gordlevsky). Only the English-written books of
Mine. Halide Edip have been conveniently accessible to the English
readers, and now M. Saussey’s selections in French will render familiar
to them some thirty other authors, among whom figures Kamal
Atatiirk himself. Each item is accompanied by very valuable
biographical and bibliographical notes.
V. M.
SEYHi DivAxixi TETKiK. Part I. By Dr. Ali Nihat. pp. 230.
Istanbul, 1934.
Shevkhi. who died pre.siimably in 829, 1425-C, is an important
figure in Ottoman literature, see E. J. W. Gibb, //Gtory of Ottoman
Poetry, i, 299-335. and J. Deny, “ Shaikh! ", in El. His chief title to
celebrity seems to be his “ rifacimento " of Nizami’s Khusrau-va-
SJilrln : much less known is his dlmn which E. J. W. Gibb, o.c., i,
305, had no possibility to consult. This divan is now the subject of
a most detailed study by Dr. Ali Nihat of the Istanbul University,
a pupil of Professor M. F. Koprtilu-zade. The author says that he
reserves the publication of the original divan with all the details on
Shevkhl’s career, and it is a matter of some inconvenience for the
readers of vol. i that vol. ii is not simultaneously before their eyes.
In the present volume Dr. Ali Nihat studies Sheykhl’s rather
vague .s«/7 and philosophical views as well as the reflections on his
260
REVIEWS OF BOOKS
works of the social milieu in which he lived (pp. 7-60). The essential
part of the book is the most detailed analysis of Sheyklil's poetical
art, which latter is. of course, chiefly traditional. The author gives
an exhaustive index of all the metaphors found in the dlnln, e.g.
under the word ahnl (evebrows) the following simile.s and epithets :
crescent moon, screen (saynhan). niihnib, letter j. crooked, fold (chin).
bow, " calamitv. " " temptation. " doorkeeper {hdjih). Each item is
accompanied by quotations from SheykhI and other Persian and
Turkish poets (Hafiz. Salman-iSavajI, Khwaju. and many others)
bearing witness to the author’s extensive and systematic readings.
Follows a special enumeration of metaphors referring to nature
(spring, morning, spring cloud, etc.). Very complete is the study of the
combined expressions (mura<it-i nazir) belonging to the same field,
e.g. the religious terms : qdmat. snjdo, qndd, nnntaz. niyyat appearing
in the same verse. A long list of Sheykhl’s de and alliteration.s
is given on pp. 204-212 : dard durd. ejmm dpseiu, etc. Finally are
quoted : Shaykhl's views on poetry ; literary forms and metres used
by him ; single verses imitated by him : his poetical licences.
This formal analysis (Formalkritik) of the works of a single author
is the necessary condition for any generalizations in the field of Oriental
poetry. It is a vain task to appreciate old Muslim poets according to
our present-day likes and dislikes. Our judgment will be put on much
surer ground when patient investigators like Dr. Ali Nihat introduce
us into the arcana^ of the old poetical workshops. In his further work
the author will find many useful .suggestions in Krachkovsky’s edition
of Ibn al-Wa'wa". Ri'pka'.s study of Baqi, and H. Ritter's penetrating
essav on Xizanu’s Bildrr.sprachr.
V. M.
Tfukest-W Solo. One woman's expedition frotn the Tien-Shan to
the Kizil-Kuin. By Ella K. Maillart. Translated from the
French by J. Rodker. pp. .407. Putnam, 19.41. lOv. 6d.
Mile. Muillart's journey consi.sted of two parts. From Moscow
she proceeded to the Republic of the Kirghiz (previously called Qarn-
Qinihiz). comprising the southern j)art of the former ” Seven rivers ”
province (Russian Setnirechye, Turkish Jiti-su). Over the Kirghiz
capital Frunze (former Piahpel) and Tokrnak she travelled to the
Is.sik-kul lake. The latter was cro.s.sed by boat to Kara-kol (former
PrzJievahl) on the south -ea.stern coast. Thence with a party of Ru.ssian
excursionists she visited the mountainous knot at the sources of the
AN EASTERN ODYSSEY
261
Narin (upper Jaxartes) and the neighbourhood of the Jangart pass in
the T’ien-shan on the Soviet-Chinese frontier. Having returned bv
the eastern hank of the Issik-kul she came to Alma-Ata (former
Ternly) the present capital of Kazakstan (republic of the " Qazaqs”,
by which term the Turkish people is meant, formerly called Qiryhiz-
Qazak).
Here began the second part of the journey. From Alma-.\ta,
situated on the Turksib " (Turkestan-Siheria Kailway) she travelled
westwards to Turkestan (Tashkent. Samarkand. Bukhara). From
Charjuy. on board a steamer, she descended the Oxus to Kharazm
(former Khiva) and, having missed the last boat plying on the Aral
sea. had to travel in winter on camel-back across the Kizil-Kum sands
to Kazalinsk (on the Tashkent-Orenburg railway).
Mile. Maillart must be congratulated on her extraordinary physical
resistance in covering all this space in the course of one autunui,
travelling third class, eating whatever food she could find in native
chay-khdnas, and sleeping in her bag in crowded stations and dreary
deserts. Still more astonishing is the confidence which the citizen of the
Swiss republic which has not even recognized the Soviet Government
must have inspired in JIoscow to enable her to pass ” without let or
hindrance “ through the remote regions of Central Asia. As the author
herself points out (pp. 15, 26. 27, 30. 139, 166, 219, and 262) that she
did not feel herself bound by this confidence, her good luck in bringing
her journey to a happy end is still more remarkable.
The journey went on with a cinematographic rapidity, yet
Mile. Maillart has a keen eye and a gift for summing up her impressions.
The book is verv readable and the numerous photographs are quite
successful.
Y. 51.
An E.\stern Odyssey. The Third Expedition of Haardt and
Audouin-Dubreuil. By Georges Le Fevre. Translated and
adapted by 51ajor-General Sir E. D. Swintun. pp. 368.
5'. Gollancz, Ltd., 1935. 18<.
Xo need to speak here of the well-known e.xpedition which the
enterprising French industrialist. A. Citroen, .sent in 1932 to connect
Beirut with Peking by land on motor-vehicles equipped with
caterpillar-tracks. As 51. G. Le Fe\Te pleasantly puts it : “ Having
... to some extent conquered Time. (51. Citroen) determined to attempt
the difficult task of ma.stering Space, being convinced of the many
262
REVIEWS OF BOOKS
I'Cnefits which might be conferred on the human race i'v our intimate
and active association of Industry, Science, and Art, and by the
fostering of the mutual understanding of the peoples of the world
through the agency of journeys made across the different countries."
The agreeable trait of the book is that the triad worshipjied by the
late M. Citroen left no place to any political suggestions or insinuations
so usual in the post-war geographical literature. The most interesting
part of the hook is the description of the spectacular crossing of the
Himalava and Karakoram ranges into Chinese Turkestan and the
adventures of the two parties, the one coming from India and the other
from Peking, before they finally met near A(|su. The cars, among
frightful difficulties, were taken as far as Gilgit, but the journey
between the latter place and the Chinese territory could be carried
out only on foot and on honscback with the time-hallowed aid of
native carriers.
V. M.
History of E.\pi/jr.\tion fro.m the Earliest Times to the
Present Day, By Sir Percy Sykes, pp. 374. G. Routledge and
Sons, 1934. 25o)'.
In view of the tremendous mass of materials and books on the
fascinating subject chosen this time by Sir P. Sykes, the chief problem
for him was certainly the selection of the facts illustrating the march
of exploration in the various regions of our planet and the co-ordination
of single chapters. This object has been fully achieved within the
space available. The book is destined for general readers, but even
a specialist in a given field will be glad to refresh his knowledge of the
situation in the neighbouring domains. The te.xt is profusely illustrated
by 35 maps (many of which are borrowed from J. N. K. Baker’s History
of Geoyraphicnl Discovery and Exploratimi) and 25 plates reproducing
rare landscapes and portraits.
Any criticisms as to the absence of some important names and as
to the proportionate length of the items must certainly have in view
that in an epitome containing only 338 pages of text sacrifices were
inevitable and a mere completeness ” would have resulted in an
unreadable list of names. However, for the benefit of the subsequent
editions it may be said that the author, using chiefly the sources
available in English, may have involuntarily omitted some names
essential in the advancement of our geographical knowledge. To speak
only of Asia, it is a pity not to find mentioned in the text Richthofen's
INDIA, MINTO AND MORLEY
263
works on China, the excellent memoirs on Persia by the clever
Napoleonic officers (Dupre, etc.), Chikhachev's detailed exploration
of Asia Minor, Dubeux de Montpereux's travels in Transcaucasia,
etc. As regards the British names, H. F. B. L^mch's title to our
gratitude is not his ascent of the Ararat (p. 267), but his important
investigations in Turkish xkrmenia of which the natural complement
were Sir M, Sykes's numerous expeditions in the regions adjoining
Armenia.
p. 16. By a slip Bactria is identified with the present-day
Badakhshan. p. 46, the Chinese Fii-lin as now admitted (Blake.
Pelliot, H. H. Schaeder) has nothing to do with ttoXlv, but reproduces
the very name of Rome (hRom > fRom). p. 246, read Schlagintiveit
(instead of Schlaginwert).
V. Minorsky.
India, Minto and Morley. By Mary Countess of Minto.
Macmillan, 1934, 21s. net.
In this admirable volume Lady Minto completely succeeds in her
object. Lord Minto’s achievement in India has been belittled, in part
by the utterances of hasty politicians eager to make out their case for
modifying the reforms associated with his name, in part by the skilfully
selected Recollections of his colleague at the India Office, suggesting,
doubtless without malice, that whatever good was done in India
flowed from Whitehall and Westminster. Mr. Buchan, in writing
^linto’s biography, had an excellent chance of setting matters in their
true light. But this he failed to take. His volume, like ever^dhing
that has come from Mr. Buchan's pen, makes easy and interesting
reading. But he was too unfamiliar with Indian conditions to do justice
to his hero in the last, and by far the greatest, phase of his career.
Minto will probably go down to history as the best Governor-
General of his generation. Sir Harcourt Butler once applied to him
a striking phase. Minto was like an elephant, he said, too sagacious
to rest his weight on rotten ground. Without the brilliant qualities
of his predecessor, Minto possessed the rare faculty of discerning the
essential facts of a political situation amid all the shams and
exaggerations of political utterances. His letters, of which Mr. Buchan
made such sparing use, here demonstrate the sane, cool judgment and
the fine temper of the man. Perhaps his masterpiece was his manage-
ment of the fickle, irritable Secretary of State. Let anyone read the
present volume and Lord Morley’s Recollections side by side, and he
•264
REVIEWS OF BOOKS
will undoubtedly find his estimate of Morley's character falling, his
estimate of Minto's rising. How is a harassed Governor-General to
deal with a Secretary of State who first peevishly complains because
the resignation of Sir Bampfylde Fuller i,s not to be ascribed to his
own personal action and then proposes to appoint Fuller to the
Council of India 1 Hinto's letters show how it should be done. But
his management of Indian affairs was no less skilful, and incomparably
more important. He did not, indeed, still Indian discontent ; but
xmquestionablv he allayed it, and left to his successor a task far easier
than that which he himself had inherited. Again and again the reader
will be struck by the accuracy of his forecasts, no matter whether he
is dealing with the probable results of refusing to consult Habib-ul-
lah before the Anglo-Ru.s.sian Convention was signed or whether he is
estimating the precise moment at which the deportees ought to be
released. Everyone in the least degree interested in the twentieth
centurv history of India should study Lady Minto’s volume.
H. D.
Rise and Fulfilment of British Rule in India. By Edward
Thompson and G. T. Garrett. Macmillan, 1934. 21.5. net.
This volume aims at giving a broad survey of British relations with
India from 1600 to the present day. It is well written. It has all the
advantages of presenting from a single standpoint a consistent view
of a long and often tangled series of events. The technicalities, the
hard names, the unfamiliar terms, which repel the general reader, are
as noticeable by their absence as they are in the famous e.ssays of
ilacaulay. The book is easy to read ; it is full of interest ; frequently
it is not unfair. But it is meant for the general reader, not for the
student or scholar ; and it generally conveys the impre.ssion that its
joint authors formed their opinions first and .supported them bv
selective .study afterwards. The book, in fact, is a bulky political
pamphlet rather than an historical study. A few e.vamples will show
how hastily the work was done. It is said that the early factors at
Surat got into trouble by importing coral from the Red Sea ; their
offence lay in importing a finer quality of coral from the 3Iediterranean.
A vague recollection of the difficulties that arose between King's
and Company’s officers impels the authors to date them from 1752, and
make officers with King's commissions jealous of Clive's activity and
success. At that date not even Jlajor Lawrence held a Kind's
commi.ssion. Mr. Thompson and Mr. Garrett think Kora and Allahabad
India’s social heritage
265
were occupied by the Marathas in 1773. These districts were assigned
to the Marathas by the Emperor, but the English held them till
Hastings sold them to the Nawab Wazir of Oudh. The second Mysore
M ar is made to open at least two years before Hvder ravaged the
Carnatic. And so on. A multitude of little mistakes which might
easily have been avoided by more serious methods of work characterize
these pages. They do not profess to be based on other than printed
documents. It could not well be otherwise, and offers no grounds of
reproach. But it is a singular thing to find in the account of the
Amir Abdur-Eahman. sandwiched in between references to published
matter, a solitary reference to the Foreign Office scries of Central Asian
Papers, at the Pubhc Eecord Office. Have the authors worked through
that voluminous and valuable collection ? If they have, it seems strange
that they found nothing else worth quoting ; if they have not it would
have been more in accordance with historical ethics to cite the
monograph which, as far as I can judge, they merely paraphrase.
H. D.
India’s Soci al Heritage. By B. S. S. O’Malley. Clarendon Press,
1931. 5s. net.
This little volume has an interest quite unrelated to its size.
Mr. O’Malley has already written a peculiarly well-informed volume
on caste ; here he deals not only with caste, but also with other
Indian social institutions and practices — the tribes which still survive,
the village community, the Hindu family, the marriage system, and
the like. These are matters of the greatest interest, especially at the
present time, when the main problem of India is how to reconcile
a modern political system with an ancient social organization.
Mr. O'Malley’s Avork may be warmly recommended to all who desire
a summary description of topics generally treated, when treated at
all, in elaborate and technical legal works.
H, D.
Anquetil-Duperron. Sa Vie. par Raymond Schw.ab. Usages des
Parses. Deux cssais du Dr. J. J. Modi. Paris, Leroux, 1931.
This interesting little volume, publi.shed ivith the aid of the Parsi
community and others, falls into three separate parts. One consists
of an abridged translation of parts of Dr. Modi’s volume published in
1916 on Anquetil-Duperron’s relations with Dastur Darab. Another
consists of Anquetil-Duperron’s essay on Par.si customs. The third,
266
REVIEWS OF BOOKS
and principal part, comprises a sketch of Anquetil-Duperron’s
singular career. To English readers this adventurous French hero
is familiar enough as the foimder of Iranian studies in the West, and
as the object of an unmerited attack by Sir Wilham Jones. But few,
perhaps, are acquainted with the life he led at Pondicherry, in Bengal,
and at Surat, devoted to the pursuit of learning under the shadow of
those repeated misfortunes which in 1761 drove the French for a while
altogether from India and left Pondicherry a heap of ruins. M. Schwab’s
chapters convey a lively impression of the headlong, reckless zeal of
his hero, and the marvel is that a man of such a combative temper
could possibly have steered clear of the warfare going on all round him.
In fact, he seldom took any part in the great historical ev'ents of his
day. He withdrew from Chandernagore before Chve captured it ; he
lived but for short periods at Pondicherry ; and the place of his most
prolonged residence was Surat, where after a while he found himself
under English protection. It was as well. He could have done nothing
to stem the tide of disaster ; and his brilhant gifts would have been
wasted in disastrous battles or futile intrigue. M. Schwab shows us
how clearly and devotedly he set before himself, as his prime motive,
the duty of acquiring eastern languages, and above all the language
and the manuscripts in which Zoroa.strian doctrine was enshrined.
Xone the less, we get striking glimpses of men and events — of Legrit’s
tall thin figure, crowned with the high starched cap with which our
ancestors replaced the hot and heavy wig on all but ceremonial
occasions ; or of that perilous journey from Kasimbazaar southwards
to Bussy's camp, in the course of which he had to sell his sword and
equip himself with bow and arrows. The vngour of his constitution
must havm been immense to endure the repeated diseases and wounds
that befell him. It is, indeed, often the case that genius is found allied
with a specially tough body. So it clearly was with Anquetil-Duperron.
H. D.
The CoSTiXEKT or Asi.v. By L. W. Lyde, Macmillan, 1933. 16s. net.
This admirable vmlume, with its wealth of sketch-maps, fills a
long-felt want. In no region of the world have geographical factors
played a greater part in limiting and directing human developments
than thay have in Asia. Xowhere should the joint studies of history
and geography be more closely linked together. This survey by so
excellent a geographer as Mr. Lyde should he in the hands of all who
are studying the history of any part of Asia. To some extent it provmkes
THE FIRST CENTURY OF BRITISH JUSTICE IN INDIA
267
a comparison ivith Dr. L. D. Stamp’s work covering the same ground.
It seems to the present writer that the latter might well be more
useful to the economist, and the former to the historian. Mr. Lvde
is mainly concerned with the geographical factors, rather than the
economic uses to which they have been put in modern times ; and
since those factors, with certain exceptions, have been relatively
constant, his pages are as interesting to the student of the past as to
the student of the present. Mr. Lyde's work is one which may be used
with all confidence.
H. Dodwell.
The First Century of British Justice in Indl\. By Sir Charles
Fawcett. 8vo, pp. xx and 269. Oxford : Clarendon Press :
London : Humphrey Milford. 1934. 15s.
This is the authoritative work of a patient and careful scholar.
Sir Charles Fawcett has had the advantage of a lifetime spent in the
marshalling and unravelling of evidence ; and it is hardly to be
wondered at that he occasionally utters a dry and well-deserved
censure on the “ picturesque writing " of his predecessors in the same
field : see, e.g., pp. 110-11. In addition to its principal theme, the
relations of executive and judiciary, the work raises many other
questions of interest, and if the answer to those questions depends
upon evidence which has not been discovered by Sir Charles Fawcett,
we must regretfully conclude that the evidence is no longer in existence.
One would like to know, for instance, who were the legal advisers
of the Court of Directors when the Company's Laws of 12th February,
1669, were drawn up. They are a noble document tinged with the very
best kind of tolerant Puritanism and far in advance of stay-at-
home English legal conceptions of the time. In part no doubt they may
be explained by local conditions and by the circumstances of a trading
company holding a delegated sovereignty (compulsory registration of
deeds, for example, figures also in the laws of other chartered com-
panies of the time) : but the severity of indigenous and Portuguese
penal laws was as great as that of England ; and it is amusing to find
the plea, familiar to the ears of a modern magistrate iir the words
■‘pith par maro, pet par mat maro ".i solemnly put forward in the
reign of Charles II as an argument for leniency.
'■ Beat me on my l)aek not on ray stomach." i.e. Whip me. but don't put me
on short commons by fining me. See pp. 70-1, where Judge \\ ilcox quotes what was
obviously this plea as an argument against corjioral punishment.
268
REVIEWS OF BOOKS
It would be interesting, too. to have the exact words of the reference
to caste customs in the Proclamation of 1718 described on p. 173 ; for
the contrast in the attitude of the Courts to Hindu law and
Muhammadan law respectively is instructive. The Moormen, as the
Muslims were called, were a single communitv with an established
judiciary and legal system ; and the appointment of their Qazi in
1694 (p. 184) to be “chief judge and decider" among them merely
recognized existing fact. In the same spirit we find that one of his
successors in office, being the ob\’ious leader of the Muhammadan
community, was appointed to be the first Muhammadan justice of
the peace. The position as regards Hindus was different. No
professional Hindu lawyer or judge had any recognition as such :
Hindu cases were referred in the first instance, not to the Shastris
(of whom there is never a word), but to the heads of the ca.ste concerned.
Even the representative Hindu chosen in 1717 (p. 171) to be the first
Hindu justice of the peace was not a Brahman nor even of twice-
born caste, but a Sudra merchant ; and the only emplojunent of a
Brahman as such in judicial procedure appears to have been (p. 186)
in the priestly function of administering an oath. The Court attempted
to control the caste jurisdiction to an extent which would not be
attempted to-day (see p. 219) ; but in the prominence which it gave
to caste panchayats it was probably more in accord with Hindu
orthodoxy than the system which prevailed from 1774/1798 to 1861
of having learned Brahmans as the sole channel through which the
English Courts obtained advice on Hindu law. According to the
Shastras themselves and according to what is known of the ancient
practice of Hindu kingdoms, the King's Court, where he sat with his
Brahman advisers, was a court of last resort : not only so, but its
law was a law of last resort, though colouring to an extent which will
always be matter of controversy the subordinate laws of the castes.
In this connection it i.s tantalizing that we can never have more
details of the “ Gentue will mentioned on p. 199. Sir Charles suggests
that this may have been merely a division of self-acquired propertv
by the father ; even .so, it would appear to cast some doubt
on commonly accepted theories of the origin of Hindu wills, since it
has been generally held that the Shastric power of a Hindu father to
make an unequal partition between his sons could only be exerci.sed
in his lifetime. If the case goes further than this and reallv means
that the Bengal vnew of the powers of a father over ance.stral propertv
wa.s an arguable proposition in Bombay in 1724, the commonly accepted
THE VYAVAHIEAMAYCKHA OF XILAKAXTHA 269
view of the authority of the Mitakshara would have to l)e revised.
Perhaps it is as well that we can never know more of this case !
One last word or plea ; the author twice speaks (pp. xviii and 209)
of the “ Black Hole “ as if it were a certain historical fact. This is
a question on which it would be of the hiwhe.st value to have the
judgment of a calm judicial mind, especially from such a thorough
investigator as Sir Charles Fawcett. To the pre.sent reviewer, at least,
it has always seemed that the late Mr. Little, when he made his
celebrated attack on the story in the pages of Bengal Past and Present,
spoiled a fundamentally good case bv the e.'ctravagance of his
advocacy and his obvious desire to trail a coat " and invite all and
sundry (culminating in Lord Curzon) to " step on it ". This much at
least is certain : Dr. S. C. Hill, a careful scholar, had already reduced
the fantastic total of deaths ; Holwell. on who.se unsupported evidence
the tale appears to rest, was an essentially untruthful person : the
tale itself is full of improbabilities. And, if the result of a dispa.ssionate
judgment should be to banish the story once and for all from the pages
of reputable history, we need not regret its passing. It is far more
discreditable to the Englishmen than it is to the Indians concerned.
The latter were guilty at worst of mere callous stupidity ; the former
are represented as giving up Calcutta with hardly a blow and allowing
themselves to be herded into a pen like a flock of .sheep. Is it an
argument for the truth of the story that the defender of Arcot was
outspoken in his contempt for the ” gentlemen of Bengal ? ''
S. V. FG.
The Vyavah.aramayukha of Nilakaxtha. Translated into English
with e.xplanatory notes and references to decided case.s by P. V.
K.A.XE and S. G. Patwardhax. 10 x 6|. pp. xx.xi + 307.
Bombay, 1933. Es. 7.
The MagfikJias or ” Rays ” are an encyclopedia of Hindu learning
compiled earlv in the seventeenth century, of which the volume on
Vyavahdra (justice and law) has become in We.stern India a work of
great and. in some places, paramount authority. It has already been
translated into English three times : but. e.xcellent though Eao Saheb
Mandlik's work in particular was. the present editors con.sidered that
there was room for a new translation in the light of recent knowledge
and including those portions of a merely academic intere.st (e.g.
ordeals) which their predecessors had omitted. [Incidentally one is
glad to observe their tribute to the Eao Saheb in the preface : without
270
REVIEWS OF BOOKS
it some of their criticisms might have been read more severely than
thev probablv intended.] They have added valuable footnotes, giving
reference to decided cases, with criticisms where necessary. The
resulting work ought to be of great value not only to scholars but to
practising lawyers and judges in the Bombay Presidency and Berar.
Messrs. Kane and Patwardhan do not express their views with quite
so much subjectivity and vigour as Rao Saheb Mandlik ; but one
gathers (p. 113) that they agree with him and, indeed, with every
scholar of repute as well as with a strong majority in the lower Court
in thinking that Lord Hobhouse's judgment in Bhagwansingh r.
Bhagwansingh (1899), 26, i, A. 153, was mistaken. Another point of
some interest is in their discussion of bandhu inheritance (pp. 164-5) ;
it is refreshing to hear a conservative and orthodox Hindu calling for
legislation to straighten out a tangle. The discussion would, however,
be more valuable if the learned editors had given us their own views :
and they hardly seem to realize the extent to which the tangle has
been cleared up by the decisions in Jotindra v. Xagendra (1931),
58, i, A. 372, and Gajadhar r. Gaurishankar, 54, All. 698, F. B. (both
of which they quote), and Mahant Garuddas v. Mahant Laldas (1933),
60, i, A. 189 (which they do not quote). It is, in fact, perfectly possible
now to state the law on Sarvadhikari's principles in an intelligible
form, which Sarvadhikari himself never succeeded in doing. There
would still remain the question whether he was right, and on this
point Kesar Singh v. Secretary of State, 49 M., 652 (quoted on p. 117
in another connection), should also have been referred to.
Good, however, though this book is, it is impossilile to welcome it
without a tinge of regret. When he brought out the first volume of
his HiMorij of Dharmmastra in 1930, Mr. Kane held out hopes of a
second volume dealing with the development of legal ideas in the
•sastric literature. Anything which distracts him from that great
task is to be deprecated. How necessary it is may be seen from one
example. On p. 105 Mr. Kane says : " Only those who have made
a deep study of the Mlriiah.sa can follow the discussion here,’’ and
similar remarks and brief discussions are plentiful. Now of this Purva-
Mlman.sa school of philosophy. Macdonnell wrote : “It lays special
stress on the proposition that articulate sounds are eternal, and on the
consequent doctrine that the connection of a word with its sense is
not due to convention, but is by nature inherent in the word itself.
Owing to its lack of philosophical interest, the system has not as vet
much occupied the attention of European scholars.'’ But whatever
THE PHONETIC SYSTEM OF ANCIENT JAPANESE
271
may be its repulsion for tlie philosopher, for the lawyer it is supremely
important. It established a canon of legal interpretation as strict as
and at some points not very different from that which English lawyers
apply to Acts of Parhament. Xearh' all the later Sanskrit law writers
of importance are saturated in it ; and a clear legal exposition of its
methods, which Mr. Kane could give us, would make the understanding
of their works a great deal easier.
We have, of course, Ganganatha Jha's translation of the sutras
of Jaimimi ; but this does not illustrate the legal workings of the
system as Mr. Kane could do.
S. V. FitzGerald.
The Phonetic Sa'stem of Ancient Japanese. By S. Yoshitake.
James G. Forlong Fund, Vol. XII. 8f X 5f, pp. xii + 71.
London ; The Royal Asiatic Society. 1934. 6s.
In September, 1931, Professor S. Hashimoto published an article
entitled Jodai no Bunken ni sonstiru Tohtshu no Kanadzitkai to loji
no Goho, which made an epoch in the history of the phonetic study of
Ancient Japanese. The world of linguistics in Japan has since been
animated in this field of research in an unprecedented way. Articles
have been written on the phonetic condition of the eighth-century
Japanese by Messrs. Y. Endo, M. Mochidzuki, T. Ikegami, T. Miyake,
K. Yasuda, and K. Xagata ; I myself have also expressed my opinion
on the subject in short essays since 1931. Naturally it gave me great
pleasure to learn of Mr. Yoshitake 's recent publication in English of
The Phonetic System of Ancient Japanese. Not only can we, through
this volume, become acquainted with many original and instructive
views of the author, but the book has presented to the scholastic
world of Europe a part of the problems with which Japanese linguists
have been confronted, and of their achievements, neither being familiar
abroad on account of the singularity of the Japanese language and
letters.
It may be admitted as justifiable that, for the original Chinese
sounds of the characters used as the Man-yo-gana, the author utilized
the results of investigation carried out by Professor Karlgren. This is
because Professor Karlgren's work, in respect of the abundance and
authenticity of the materials used and the comparative completeness
of his method of study, can be considered probably as the best of all
the theories so far advanced on the same subjects by the scholars of
EE^^EWS or BOOKS
272
the East and of the West. It is only to be regretted that the author
has occasionally followed the errors committed by Professor Karlgren,
althou,(;h I shall not go deep into this ciuestion in the present review.
Nor can I say that I am in entire agreement with the author on every
conclusion of his. But, as I am limited in space, I shall note only one
or two points that have occurred to me on the rpiestion of the general
method of study.
With regard to the eighth-century Japanese syllables answering
to the modern Ixi. hi. lu. ke. ko, (/a. gi. gu. ge, and go. the author states
as follows : " The initial consonants of these .syllables are represented
bv Ach. k-. k'-'x -• !J -■ T'- This plainly shows that the Japanese
consonants did not correspond exactly to any of the Chinese sounds,
but that thev were feebly aspirated varieties [C] and [g^ as heard in
modern Tokvo dialect. It mu.st, however, be remembered that Ancient
Chinese did not pos.sess an una.spirated //. and therefore tlie Japanese
would not have been able to indicate their [//] adequately, even if it
had been an unaspirated con.sonant ■' (p. 29).
However, it seems to me rather dangerou.s to make such an a.ssertion
on the basis of these data alone. According to the results of my
inve.stigation the syllabic corresponding to the modern hi is repre-
.sentedby f.f. Sp;. :^p 7. Bf, si>|. fnj. \^, ;)jp, -to,
Si- il" in eighth-century literature, and the syllable
answering to the modern ga by |1^, fg, fnj. . f^. and
. This .shows that oidy the three characters ^5- M- •'itid tnj, are
used for both syllables, while the remaining twenty-three are
distinguished according as they stand for the modern kn or gu. In
the Kojiki occurs only in places where we should find ga in modern
Japanese, wlicreas in the Nihongi and the Man-v6-.shu it represents
the syllables corresponding to the modern ka and gu. The character
fpj is found in the Kojiki and the Man-vo-shu as answering to the
modern //u. but in the Nihongi it occurs but once as a correspondence
of the modern ku. The character fnj, which does not present it.self in
the Kojiki, is u.sed in the Nihongi only in the name of the province
^ M {huruga in modern Japanese). In the Jlan-yo-shu it corresponds
to the modern ka and ga. but the instances are not many. Each of
these three characters belongeil to the [ig group (y-) in Chinese, and
has Kan-on kn and Go-on ga in Sino-Japanc.se, according to their
modern pronunciation. It may therefore be conjectured that the.se
characters with their Kan-on value were used as a correspondence of
the modern ka. and with their Go-on value as a correspondence of the
THE PHONETIC SYSTEM OF ANCIENT JAPANESE
273
modern ga. Although Mr. Yoshitake discusses the question of the
consonant answering to the modern g conjointly with that of the con-
sonant corresponding to the modern k. it is necessary to examine them
separately if we are to accomplish our study more properly. To
consider the matter still more strictly the Man-yo-gana are the letters
that represent syllables, not single sounds. As a consequence, a mere
glance at the letters does not enlighten us as to which two syllables
had the same initial sound in the eighth century Japanese. This means
that all the syllables ka, ki^, ki^, ku, ke^, ke^, koj^, ko^, ga, gi^, gi^,
gii, ge^, ge^, go^, and go^ have to be investigated severally. I shall
here narrow down the problem, and, after first considering the nature
of the eighth-century Japanese syllable answering to the modern ka,
I shall enter upon the discussion of the syllable corresponding to the
modern ga.
For indicating the eighth-century Japanese syllable answering to
the modern ka the following Man-yo-gana are used : If , jSJ,
M, li?, j®. S. # (all of the ^ group, Ach. k -) ; of,
(both of the % group, Ach. k '-) ; §0 (of the group. Ach. g '-) ;
IST; § (both of the group, Ach. x*) ; ^ (all of the group.
Ach. y-). The only conclusion we can safely draw from this evidence
is that the initial sound of the syllable in question would have been
a velar consonant of some kind. In order to determine the precise
nature of the consonant, further facts are needed. Now the Chinese
of the early seventh century used the characters ^ (Ach. ka) and If
(Ach. ka) for representing the syllable which corresponds to the modern
Japanese A’u (Ph # ® fli)- The Japanese priest Ennin ([H fl.
A.D. 794-864), who, while studying in China, was taught the reading of
Sanskrit letters by the Indian priest Hogetsu ('^ ^ ), recorded what
he had learned in his own account of his sojourn in China. In this
the Sanskrit ka is explained as ^ ^ ;^ (“ Pronounce
it as the Japanese sound of jfjp ”) and the Sanskrit /w as jjl ^ ^ ^
{“ Pronounce it approximately as the Chinese sound
of ^ "). When we take into consideration all the points disclosed
by these data we can perceive that the eighth-century Japanese syllable
corresponding to the modern ka had likewise [A'] as its initial sound.
Moreover, from what is shown by the before-cited data, we may infer
that the sound [!•] under discussion was not a distinctly aspirated
consonant like the Sanskrit kh or the initial sound of the ^ group
(Ach. k'-) in Chinese, but that it was nearer an unaspirated consonant.
Even if it were an aspirated consonant we must consider the aspiration
VOL. \JII. PART 1. Is
274
REVIEWS OP BOOKS
to have been very feeble. In the phonetic system of Ancient Japanese
[A’] and [A'] were not distinguished and therefore the contemporary
Japanese reproduced the Chinese syllables [Art], [Aa], and [A'u] in one
and the same way [A-]. It is not yet clear whether the vowel contained
in this syllable was [u] or [a]. In the northern Chinese dialect (or
dialects) of the T ang period [^'-] and [y-] seem to have been gradually
losing their voicing, thus approaching the voiceless [A], [A ], [cj]. [cf ]. [y],
and [r] as found in modern Mandarin (cf. Karlgren : Etudes sur la
phonologic chinoise, pp. 565-9). On the other hand, Ancient Japanese
had neither [y] nor [A], The ancient Japanese therefore replaced the
Chinese syllables like [ya], [ya]. and [g'ia] also by the Japanese syllable
[A-]. This leads us to think that the Sino-Japanese sounds for ^
(Ach. Aa), Jd (Ach. Aa), hf (Ach. k'a), (Ach. g'la), |i^ (Ach. yo),
^ (Ach. ya), etc., each a,ssumed the form of [A-] at the time. It is
no wonder then that all these characters should have been used
indiscriminately as the Man-yo-gana for representing the eighth-century
Japanese syllable [A-].
For indicating the eighth-century Japanese syllable corresponding
to the modern ga the following Man-yo-gana are used : ffe.
1 ®, ® (all of the @ group, Ach. y- > yg-) ; fpj, ^ (all of
the [5 group, Ach. y-). The only conclusion we can safely draw from
this evidence is that the initial sound of the syllable, under consideration
would have been a voiced velar consonant of some kind. In order to
determine more precisely the nature of the consonant, further facts
are required. Now we find in the modern dialects of central and
eastern Japan (including Tokyo, Kyoto, etc.) the .syllable which is
repre.sented by (.spelt ya in Romaji) is pronounced [y-] when standing
at the beginning of a word and [y-] when occurring in other
positions. But in the Kyoto dialect of the sixteenth century
(as in the case of modern dialects of western Japan) jy' was
pronounced [y-] in all positions (cf. Kutshi/an Kgogi no Yoga ni (suite,
contained in Professor Ha.shimoto's Bunrokn Gnnnen Amakusahan
Kirishitan Kgogi no /iV«Ay!7.pp.35-6). Indeed, dialectallyor individually
the form [y-] seems also to have been in existence since olden times,
because Ennin ([g] describes the Sanskrit ga as
(■‘ The Japane.se .sound of the character ") while explaining the
Sanskrit ha as ^ n ± W C Pronounce it as the
.lapanese nasal sound of the character ^ "). This bring.s to light
the phonetic condition of the Japanc.se language of the first half of the
ninth century. We may therefore consider that the phonetic condition of
LES NOTES DE CHEVET DE SEI SHONAGON’
275
the eighth-century Japanese would not have differed widely from it,
although no further particulars are as yet known to us.
In short Mr. Yoshitake's argument is in many cases based on too
weak grounds. For his reconstruction of the phonetic system of
Ancient Japanese he cites the following four materials ; (1) modern
Japanese sounds ; (2) Kana version of Go-on ; (3) Ancient Chinese
as reconstructed by Professor Karlgren ; and (4) archaic Chinese sounds
as proposed by the same Sinologist (p. 16). Of these four the last three
pro'cide, after all, means for finding out the original Chinese sounds
of the characters used as the Man-yo-gana. But the majority of the
users of the Man-yo-gana knew only the Sino-Japanese sounds of the
characters, being quite ignorant as to how they were pronounced
in China. Consequently the original Chinese sounds of the characters
used as the Man-yo-gana, in most cases, do not lend themselves to a
precise determination of the nature of the Japanese syllables repre-
sented by the characters ; they merely restrict in some measure the
extent of possibility. If tlierefore a more detailed knowledge of the
nature of the Japanese syllables is to be acquired, it is necessary to
make as much use as possible of such data as (1) the transliteration of
Japanese words in Chine.se characters as rendered by the ancient
Chinese and Koreans, (2) the accounts given by the Sanskrit scholars
of Japan since the ninth century, and (3) the phonetic condition of the
modern Japanese dialects (including the Luchuan dialects). The
Man-yo-gana are certainly extremely important material, but we must
in my opinion be on our guard lest we overestimate their value.
Arisaka Hideyo.
Les Notes de Chevet de Sei Shox.agox', Dame d'Honxeor au
Palais de Kyoto. Traduction in e.xtenso de I'ancien texte
japonais. Par Axdre Beatjard. X 6J. pp. xxii + 329.
Paris ; Librairie Orientale et Americaine G.-P. Maisonneuve,
1934. 80 frs.
Sei Shoxagox', sox Temps et .sox Oeuvre (Une Femme de Lettres
de I'Ancien Japon). Par Axdre Beaujard. 9| x pp. 377.
Paris : Librairie Orientale et Americaine G.-P. Maisonneuve,
1934. 80 frs.
These two learned works by Dr. Beaujard are beyond all praise.
Seldom is a study of a literary product accomplished with such laborious
thoroughness as by the present scholar. Not only has Dr. Beaujard
276
REVIEWS OF BOOKS
given us a complete translation of the Makura-no-Soshi in an excellent
style, but he has also handled the stOl harder task of portrajdng its
author, Sei Shonagon, in an equally admirable manner through the
medium of her immortal hook.
As a matter of fact, the subject dealt with by Dr. Beaujard is not
new. The Makura-no-Soshi has been rendered into modern Japanese
by native scholars. But these versions help us little when translating
the original text into a European language. What is usually called
the “ subject ” of a sentence, which has to be expressed in one way or
another in any European language, is at times just as difficult to detect
in the Japanese translation as in the original text. This point
is well illustrated by the opening line, Haru iva akebono, which
Professor Kaneko has modernized thus : Ham wa akebono ga koto ni
omoshiroi. This translation is clear enough to be readily apprehended
by a Japanese schoolboy 15 years of age, and yet the original expression
has given rise to varied interpretations, as may be seen from the
following : —
Aston ; “ In spring I love to watch the dawn. . . .” ^
Florenz : “ (Es ist interessant, zu beobachten) wie es sich im
Friihling . . ., die in der Morgendammerung. . . ." ^
Re von : “ Ce qui me charme, au printemps, c’est I'aurore.” ®
Beaujard : “ Au printemps, (c'est) I'aurore (que je prefere).” *
Every one of these four renderings, of which the two last are
structurally identical, is indeed charming, but they cannot all be
accurate. If one of them is right others must be incorrect. The
difference, however slight it may appear, must on no account be
regarded as trivial, for it betrays how easily the intention of the original
writer can be distorted through a lack of attention on the part of the
translator. Thus while recognizing the “ impersonal ” character of
the Makura-no-Soshi (cf. N.C., p. xi). Dr. Beaujard has introduced
■■ je ” into his translation. What Sei Shonagon here discusses is
obviously the spring itself, and not what she loves or prefers in spring.
The purport of the original must then be ; “ The spring is to be best
enjoyed at dawn " or ” The spring is most delightful at dawn ".
The beauty of the laconic utterance flam ira akebono cannot, of
* \V. G. .\ston, A Iliflonj of Japanese Literature (1930), p. 106.
^ K. Florenz, Geschiehte der japanischen Litteratur (1909), p. 224.
’ M. Revon, Antholagie de la litterature japonaise, des oriijines au A'A'« siech (1928)
p. 200.
‘ A. Beaujard, Le.s Xotes de Cherel dr Sei Shonagon (1934), p. 1.
LES NOTES DE CHEVET DE SEI SHONAGON'
277
course, be expressed in any other way, not even in the manner
Professor Kaneko has paraphrased. The exquisite terseness is partly
retained in the French translations, hut is completely lost in both the
English and the German renderings.
The modern Japanese versions with commentaries, nevertheless,
serve a useful purpose, since they assist in determining the probable
meaning of ambiguous words which are met with in the original text.
Dr. Beaujard is certainly justified in exercising his own discretion when
the native commentators are at variance. He has, for instance, adopted
Professor Kaneko's reading toku yare to “ (Nous dimes au conducteur)
d aller plus vite ” (N.C., pp. xiii, 126), in preference to Mr. Mizoguchi's
version, toku yaredo “ although the carriage made haste ”, which latter
does not seem to fit in with the context. In the case, however, of such
a vague word as kakiita, the present translator might have investigated
a little more deeply. He has translated it, in accordance with Kaneko's
interpretation, by “ Un tableau (noir) ou Ton note (ce qu’on a peur
d’oublier) ”, rejecting Mizoguchi's rendering “ A board used for
carrying things ” {N.C., pp. xiii, 230). But the meaning “ A board
used for cutting out cloth as given in both the Dai-Nihon Kokugo
Jiten (1929) and the Gensen (1929-1930), seems just as suitable, or
even better than either of the significations proposed by the two
commentators.
In view of the fact that the Makura-no-Soshi has many different
texts, whose authenticities remain to be proved, the first book of
Dr. Beaujard's can broadly be described as “ a reliable translation of
Mr. Mizoguchi’s version of the Makura-no-Soshi ’’ (cf. N.C., p. xiii,
n. 1). The reader must not imagine that every word he reads in this
version was actually WTitten by Sci Shonagon. This unfortunate
circumstance makes it the more difficult to draw out facts from the
Makura-no-Soshi. Almost all that can be inferred in the present
uncertainty is set forth by Dr. Beaujard in his second volume.
The book is divided into four parts, preceded by ” Notes
preliminaires ”, under which are e.xplained (1) the transcriptions used
by the author for Japanese. Chinese, and Sanskrit words. (2) Japanese
and Chinese proper names, and (3) matters concerning the reckoning
of time in ancient Japan. Part i is devoted to a general study of the
age in which Sei Shonagon lived. It opens with a brief account of
Japanese history from the earliest times to the tenth centurv and of
Chinese historv till the T’ang period. This is followed by '' Le Japon
au x"^ siecle ”. under which heading the author discusses (1) the world
278
REVIEWS OF BOOKS
in which Sei Shonagon lived, (2) the organization of the State, and
official titles, (3) the religions (including Shinto, Buddhism,
Confucianism, and Taoism), divination, magic, and superstitions, (4)
science, arts, and recreations, (5) contemporary thought, manners,
customs, and clothing, and (6) life at the Court.
These are but preliminary to part ii. in which a particular study is
made of the important families, touching upon the origin of Sei
Shonagon. After a careful examination of the events told in the
Mabura-no-Soshi. Dr. Beaujard ascribes her parentage to Kiyowara
Motosuke. without however, referring to the clear statement made in
the Dai-Nihonshi (a.d. 16.57-1906), where we read ; '' Sei Shonagon
was the daughter of Kiyowara Motosuke, Governor of Higo." Part ii
is concluded with a chronologically arranged list of the events quoted
in the Makura-no-Soshi.
The remaining 155 pages are the cream of the book. In part hi
Dr. Beaujard dwells on the nature of the Makura-no-Soshi. He begins
with a discussion of the circumstances in which it was written and
an e.xplanation of the commentaries, followed by a study of the
language and .style of the Makura-no-Soshi. Then he compares Sei
Shonagon's poems with those of which they are adaptations, and asserts
that her poems do not e.xcel those of her contemporaries (p. 232).
The contents of the Makura-no-Soshi are cla.ssified into eight categories
by the present author : (1) Enumerations. (2) Descriptions et tableaux,
(3) Divers (les .sons, les parfums. . . . ), (4) Recits, (5) Reflexions,
(6) Portraits physique.s et moraux. (7) Les per.sonnages — Tvpes
generaux. and (8) Per.sonnages particuliers. These, however, may
ultimately be brought under the three broader categories: (1)
Classification. (2) Go.s.sip, and (3) Reflections and oljservations, as
suggested by Profe.ssor Ikeda.
Part iv treats of .Sei .Shonagon herself. In the opinion of
Dr. Beaujard. .she was ” coh'ueuse " (p. 258) and " ingrate ” (p. 258),
but proved herself .soimUimes “ charitable " (p. 258) and " delicat*' "
(pp. 258-9); knew how to l>e thankful when she was itiformed of some-
thing she did not know (p. 259), does not seem to have had a verv good
opinion of people in general (p. 260). appreciated the a'sthetie sidi' of
things rather than their moral aspf'cts (p. 261). and was a .sinc(>re.
but not ardent, believer in Buddhism, and in the main observed tlu'
doctrines of the Tendai .school (p. 266). If these are. among others, the
inferences that may be drawn from the Makura-no-.Soshi, the more
tender side of 8oi Shonagon can be perceived in the potmis (■(dlected
LES KOTES DE CHEVET DE SEI SHOXAGOX'
279
in her Kashu "■ Recueil prive ", which is dismissed by Dr. Beaujard
in less than twenty words on p. 187.
After scrutinizing Sei Shonagon's scientific and geographical know-
ledge and her allusions to the history, legends, and literature of Japan
and China, as revealed in her work. Dr. Beaujard concludes that she
was no more learned than the people who surrounded her (p. 350).
He thus seems to agree with Mr. Waley in his judgment that “it is
her extreme readiness of wit rather than her erudition that makes
Shonagon remarkable (cf. p. 351. n. 3). In support of this
deduction may be cited many instances, of which the following is
perhaps the most famous. One winter's day, when there had been
a heavy snowfall, the Empress asked the Lady-in-waiting : " Well,
Sei Shonagon. How is the snow on the peak of Koro ? "
To this Sei Shonagon replied simply by having the lattice raised
and rolling up the bamboo screen herself in allusion to the following
poem by Po Chii-i (which was no doubt in the mind of the Empress) ; —
Le soleil est d4ja haut, et j'ai assez dormi ;
mais je paresse encore avant de me lever.
J'ai entasse (sur moi) les couvertures (aussi haut qu'une)
petite tour, et je ne crains pas le froid.
La cloche du temple d'l-ai, je I'entends en me soulevant sur mon
oreiller ;
La neige du pic de Hsiang-lou, je la vois en relevant le store de bambou
(S., p. 343).
“ II faut avouer, " Dr. Beaujard rightly observes. " que Sei ne
pouvait guere moutrer avec plus d'elegance son erudition, ou. pour
mieux dire, sa presence d'e.sprit. puisque, comme elle-meme I'ajoute
aussitot, les vers de Po Tchu-i etaient familiers a toutes ses compagnes "
(S., p. 343).
Copious notes throughout and a comprehensive bibliography at
the end of the second volume (pp. 363-374) complete the two invaluable
works under review. The study would have been brought nearer to
perfection had mention been made of the pa.ssagcs illustrative of the
life of the lower classes, as. for c.xample. the description of the men and
women engaged in the planting and reaping of rice (A.f.. pp. 24b.
249), and that of the fishcrwomen (X.C.. p. 294). But the omission can
hardly prevent us from assuring the author that he has succeeded
in bringing us to a closer contact with the brilliant figure of the tenth-
eleventh century Japan who has been compared to the Fleur de
cerisier ".
S. Yoshitake.
OBITUARY
Professor Alice Werner
The death of Dr. Alice Werner, which occurred at Welw\’n Garden
City on 9th June, 1935, was an irreparable loss not only to African
studies but to all who knew her personally. Born at Trieste on
26th June, 1859, she had lived in her early youth in New Zealand,
Mexico, the United States of America, and on the European Continent,
and was educated partly in Germany and later on in England,
eventually going to Newnham College, Cambridge. Thus she became
a good linguist, with a cosmopolitan outlook in the best sense of
the word.
But the ultimate goal of her destiny was Africa, and a visit in 1893
to Nyasaland, followed by one in 1894 to Natal, laid down the main
lines of her subsequent career. In 1899 she undertook the task of
teaching Afrikaans and Zulu in London, and another African tour in
1911-13, this time to East Africa, brought her into contact with
Swahili and other languages of that region.
In 1917 she joined the School of Oriental Studies as one of the
original members of its teaching staff ; and, rising through the grades
of Lecturer and Reader, she eventually became Professor of Swahili
and the Bantu languages. During the years that intervened till her
retirement at the end of the session 1929-1930 hundreds of students
passed through her hands and benefited by her teaching. During part
of this time she also gave instruction at Oxford and Cambridge, and
both there and in London her .sister. Miss Mary Werner, co-operated
with her.
In 1928 Dr. Alice Werner received the degree of D.Lit. (London)
and in 1930. after her retirement, she was given the title of Emeritus
Professor by the same University ; in 1931 she was awarded the silver
medal of the African Society, of which she was a Vice-President, and
also received the honour of a C.B.E.
In addition to her varied activities as a member of the teaching
staff of the School, she made a number of contributions on African
subjects to our Bidletin. as well as to other journals, and produced
several standard works of permanent value on African philology and
nn-thology. But tho.se of her colleagues who are not specialists in the
282
OBITUARY
African field, while fully recognizing her great achievements as a
scholar, will chiefly remember her as an amiable and genial friend and
feUow-worker in our School.
C. 0. B.
In Memoriam : Jarl Charpentier
There are many who grieve for the loss of the great orientalist
who passed away with startling suddenness last summer, and amongst
the sincerest mourners are his friends in this School, with which for
many years past he was associated by intimate ties. Here he delivered
a lecture of very high importance on the original home of the Indo-
Europeans, which was printed in this Bulletin (IV. 1 47 ff.) ; and our pages
were often enriched by other articles and reviews by him. In losing
him we have lost very much.
Jarl Hellen Robert Toussaint Charpentier was born in Gbteborg
on the 17th December, 1884. His father. Major Robert Charpentier,
was of French e.vtraction. and the blend of Gallic and Scandinavian
blood bore happy fruits in his son. His Alma Mater was the ancient
University of Uppsala, where he became Docent in Sanskrit and
Comparative Philology in 1908. He speedily made his mark. He
had very great powers of work, with a notably retentive and orderly
intellect, to which was added the rarer gift of a sound and vigorous
judgment. In his short life he wrote much — innumerable articles in
learned periodicals and not a few books in his native Swedish, English,
French, and German — but all that he wTote was of high quality, and
his University duly recognized his outstanding talents by promoting
him in 1927 to the chair of Sanskrit and Comparative Philology in
succe.ssion to his master Johan.sson. In purely linguistic studies his
best work was perhaps Die Demlerutivbihlutuien der indoirnnischen
Sprachen. which appeared at Upp.sala in 1912 as vol. G of the Archins
d'£tudes Orientales ; but he produced many other c.vcellent philological
studies, e.g. De indaeuropeiske Spraken (Uppsala. 1915), and Die
verbnlen r-Endnmjen der Indogermanischen Spracheti (ibid.. 1917).
He gave more attention to the ancient languages of India - Sanskrit.
Pali, and Prakrit — with their literatures, and to Indian historv and
antiquities. His Paccehabuddhugeschichlen. of which the first part was
issued in the Uppsalu V niversitcts Arsskrift of 1908. and his Buddha
(Stockholm. 1910) revealed wide knowledge of Buddhist te.xts ;
and his critical edition of the Uttarudhyaipiua, a Jain Prakrit canonical
JARL CHARPENTJEE
283
book, which he published with notes and commentary as vol. 18 of the
Archives d'i^tudes Orientates at Uppsala in 1922. bore striking evidence
to his mastery of Jainism. Between these two he published a very
valuable work on Brahmanic literature. Die Suparnasage (no. 26 of
the publications of the Vilhelm Ekinans Universitetsfond. Uppsala
and Leipzig, 1920), in which he edited with translation and full
exposition the Suparnadhyaya, a curious semi-dramatic poem of the
later Vedic period narrating the m\-th of Garuda. Another notable
contribution to the study of Hinduism is his Brahman, which was
issued in the Uppsala Universilets Arsshrift of 1932. bd. 2 : this is an
investigation into the origin of the term brahman and the religious
and mythical ideas attached to it in ancient India, including a vigorous
criticism of Professor Hertel's theories on the subject. In history and
antiquities also his pen was very active. He contributed to vol. i
of the Cambridge History of India an excellent chapter on the Jains ;
in 1924 he brought out at Stockholm a handy little Stora Moguls
Dynasti ; and his last work was an able study of the Indian travels
ascribed to Apollonius of Tyana.
No less important was his work upon the relations of the early
European travellers in the East, especially the Catholic mi.ssionaries.
Many years ago his keen eye perceived the value of these neglected
sources, and he set himself to examine them, with fruitful results.
The first of these studies .saw the light in the pages of this Bulletin
(II, 731 ff.), under the title " Preliminary Report on the ' Livro da
Seita dos Indios Orientals' (Brit. Mus. Sloane 1820)”; then came
“ A Treatise on Hindu Cosmography from the Seventeenth Century
(Brit. Mus. MS. Sloane 2748 A) ” in this Bulletin, vol. III. 317 ff..
and ” The British Museum MS. Sloane 3290. the Common Source of
Baldaeus and Dapper ". ibid.. HI. 413 ff. Father Fenicio s Ltvro da
Seita dos Indios Orientais, the subject of the above-mentioned
■■ Preliminary Report ”. was at last edited l)y Cliarpentier with
introduction and notes in 1933 at L ppsala. as no. 40 of the publications
of the ^'ilhelm Ekinans Universitetsfond.
These notices of a few of his works will give some conception of the
great and fruitful energies which were crowded into Cliarpentier s
short life. Would that he had been spared to the full span of year.s.
But in his brief time he achieved much, marvellously much, and he
leaves with us a cherished memory of a great scholar and a loyal
friend.
284
OBITUARY
Professor Basil Hail Chamberlain
It is with deep regret that we have to record the death at Geneva
on the 15th of February of Professor Basil Hall Chamberlain.
A son of the late Vice-Admiral William Chamberlain and a grand-
son of Captain Basil Hall, the navigator, after whom he was named,
he was born at Southsea on the 18th October, 1850. From childhood
he was delicate, with a weak constitution, and it was perhaps for this
reason that his earlier education in England was left in the hands of
private tutors. At a later age he was sent abroad to the Continent to
learn French and German, and when, after some years spent in French
lycees, he returned to England with his education finally completed
he was given a post in the house of Barings. But work on an office
stool proved quickly unsuitable and, his health breaking down, he
left home on a prolonged voyage. It was in the course of this that,
after various wanderings, he eventually, in 1873. reached the shores
of Japan, a country which was henceforth to be his real home for
nearly forty years.
At the moment of his arrival enthusiasm for the learning and the
things of the West was intense and general ; and young Chamberlain,
already an accomplished classical scholar and possessed of an
exceptionally intimate knowledge of French and German, had no
difficulty in obtaining an appointment as teacher of English in the
newly founded Imperial Naval Academy. A linguist by predestination,
he at once set himself to the study of Japanese ; and, mastering its
initial difficulties with exceptional ease, he quickly passed from the
stage of student to that of scholar and authority. In 1880 the first
fruits of his industry appeared in the shape of a volume on the
Classical Poetry of Japan. This was followed in 1886 by a Romanized
Japanese Reader and a Simplified Cramtnm of the Jajnnese Lanyuage,
modern written style ; then came in quick succession a Translation
of the Kojiki, with Commentary, and his Handbook of Colloquial
Japanese. The last and Aston's .slimmer and more concise work on the
same subject were for years the standard textbooks for English
students of the language. In the meanwhile, he was also a con.stant
contributor of learned papers to the Journals of the Asiatic Society of
Japan, an association founded in 1872 under foreign auspices for
purposes of research. In 1890 he published Things Japanese, a vade-
mecum of miscellaneous information which enjoyed a wide popularitv
and proved invaluable to serious student and casual globe-trotter
alike. Next he brought out. in collaboration with the late Hr. W. B.
PROFESSOR BASIL HALL CHAMBERLAIX
285
Mason, a revised and up-to-date edition of Murrays' Guide Boole to
Japan, based on the singularly comprehensive and learned work,
compiled originally by Satow and Hawes. Long before this he had
severed his connection with the Xaval Academy ; but in recognition
of his services in the field of linguistic knowledge he had been
subsequently appointed Professor of Philology in the Imperial
University of Tokyo. In 1895 he published a Grammar of the Luchuan
Language ; in 1902 he wrote for the Asiatic Society of Japan a paper
on “ Basho, and the Japanese Epigram ; and in 1905 appeared his
Moji no Shiritbe — a Practical Introduction to the Study of Japanese
Writing, a sumptuous and beautiful!}' printed work destined to lighten
the labours of generations of unfortunate students. His paper on
“ Basho and the Japanese Epigram ”, in the opinion of some the most
brilliant and perfect piece of work that ever came from his pen, was
subsequently published in book form in combination with his earlier
volume on Japanese poetry. His last contribution in the field of
Japanese knowledge was a small pamphlet entitled The Invention of a
New Religion, an arresting essay but unpalatable to many Japanese.
Long before this, however, ill-health and poor eyesight, the latter
doubtless aggravated by the strain of reading the Chinese ideograph,
had made him a semi-invalid ; and, never a frequenter of the
fashionable world in Tokyo, he now lived in practical retirement,
accessible only to his more intimate friends. Finally, in 1911, he
decided to leave the country and returned to Europe, settling
ultimately at Geneva. There, having laid aside for good the work which
had occupied him for so many years, he turned to the study of French
bterature later embodying the results of his labours in a work entitled
Eight Centuries of French Poetry. His last book appeared as late as
1933 and is called Encore est vice la Souris, a title taken from a poem
of Charles of Orleans in which the latter protested against rumours
that he w'as dead, Chamberlain having seen himself similarly referred to.
He was a great scholar and a great linguist in the true sense of
the term, a trained philologist, with a profound knowledge of the
languages of the Far East and of the West alike, gifted with amazing
powers of intellectual application, and a master of style, whether the
medium of e.xpression was French or English, gentle and retiring in
manner and address, but accessible always to the humble student in
search of advice or assistance.
Thus has left the stage one of the greatest autliorities on Japan
and the Japanese language that this country has produced.
Harold Parlett.
286
OBITUARY
Professor Zoltan Gombocz
In Dr. Zoltan Gombocz. Professor of Hungarian Linguistics in
Budapest Lniversity, Hungary, has lost one of the finest combinations
of Hungarian and Western culture.
A peculiar manv-sidedness characterized his whole scientific
activity. His researches comprised all the fields of Hungarian
linguistics — phonetics, sound history, accidence, svTitax. and word
history. The results of his labours are concentrated in two com-
prehensive works, the EtyttmJogical Dictionary of the Hungarian
Language, which, in collaboration with his friend Professor John
Melich, he began to compile in 1914. and which, though still
unfinished, is the most exhaustive dictionary of its kind, and in his
famous monograph on Bulgaro-Turkish Loan-words in Hungarian,
published in Vol. xxx of the Memoires de la Socie'te Finno-Ugrienne.
The latter work threw a new light on the prehistoric wanderings of
the Hungarians and on the main problem of Hungarian folk-lore,
the genesis of the Attila legend.
His amazing mastery of languages and his exhaustive bibliographic
knowledge enabled him to make himself familiar with all the results
of foreign methodology and to apply these results to the particular
problems of Hungarian linguistics. But he did even more than that ;
he supplemented these results with researches of his own dealt
with in his “Linguistic Methodology” (N yelvtdrteneti modszertan, 1922)
and in his Semasiology " {Jelentestan, 1926), as well as in a number of
important articles and presidential addresses. Although his w'hole —
purely psychological — philosophy of language is based upon the
functionalism of the Geneva school (Saussure, Bally, Yendryes),
he remained open to other foreign influences too. He wms, of course,
deeply interested in English linguistics in the work of phoneticians
like Miss Soame.s and Daniel Jones, and in the methodological
researches of Otto Jcsper.sen. .some of who.se views he discussed in the
very last article written by him (" Funkcionalis nyelvszemlelet
= “ Punctional A.spects of Lingui.stics ” in Magyar Xyelr, 1934).
It was in his inaugural address at the Ebtvbs College in October.
1914. that he summed up his creed when he .said that “ in the life of
every scientist there is one deci.sive moment, when the sacred
enthusiasm of the search for truth gets possession of his mind '. And
his paramount significance consists, not only in his undving .scientific
achievements, but also in that noble and generous educational activitv
of his which aroused this .sacred enthusiasm in two generations of
young Hungarian scholars. Stephex Ullm.\nn.
TITLES AND SUMMARIES OF THESES
By Students in the Department of Ancient India and Iran,
1921-1935
Studies in Sanskrit Poetics. By Sushil Kumar De (D.Lit., 1921).
Published in altered and enlarged form under title ; Studies in
the History of Sanskrit Poetics. London : Luzac and Co., vol. i,
1923; vol. ii, 1925.
The origin and development of the Bengali Langl^age. By
Suniti Kumar Chatterji (D.Lit., 1921).
Published by Calcutta University Press, 1926.
The fundamental principles of Sa.mkara-Vedanta. By
Bibhutibhushan Raychaudhuri (Ph.D., 1922).
Copy in University Library, London.
Doctrinal evolution and literary history of Sariikara-Vedanta.
^aihkara's theory of Adhyasa, Brahma. Maya, Jiva, and its relation
to Brahma and Atman. His doctrine of salvation.
The Puk.1n.as : their composition and correlation. By
Kullidaikurichy Naraimer Sitaram (Ph.D., 1922).
Copy in University Library, London.
The history of Pali literature in Ceylon. By G. P. Malalasekera
(Ph.D., 1925).
Published under title : The Pali Literature of Ceylon. (R.A.S.
Prize Publication Fund. No. 10.) London : Royal Asiatic Society, 1928.
The Divine Power : Beiii" an historical study of the conception
of the Sakti in Hindu thoutrht interpreted from the original texts.
Bv Sudhendu Kumar Das (Ph.D., 1925).
Copy in University Library. London.
Development of the idea of sn /.O' from Rgveda through Bralimanas
and Upanisads. Sahi in the Trika School of Kashmir and in the
Lihgayata School.
The Phonology of P.vnjabi as spoken about Ludhiana. By
Banarsi Das Jain (Ph.D., 1926).
Published under title ; A Phonology of Panjabi and a Ludhiana
Phonetic Reader. (Panjab L^niver.sity Oriental Publications, No. 12.)
Lahore : University of the Panjab, 1934.
288
TITLES AND SUMMARIES OF THESES
The idea of Karma and Eeincarnation in Hindu religion and
PHILOSOPHY. By P. H. Ye\-titch (Ph.D., 1926).
Copy in University Library. London.
Traces growth of the idea of Karma in Egveda and Brahmanas,
older Upanisads, early Buddhism. Mahabharata. Sariikara, Eamanuja.
and the Surya-glta. Bibliography.
The PHONOLOGICAL data of THE InDI.AN SYSTEMS OF PHONETICS.
By Siddheshwar Varma (D.Lit.. 1927).
Published under title ; Critical studies in the phonetic observations
of Indian grammarians. (James G. Forlong Fund, vol. \'ii.) London :
Eoyal Asiatic Society, 1929.
Influence of Kanarese on Marathi syntax. By Eangacharya
Yasudevacharya Jahagirdar (M.A.. 1928).
Copy in University Library, London.
Influence of a Kanarese substratum is suggested to explain certain
peculiar uses of postpositions in Jnanesvarl and some details of the
sjTitax of the old present tense and the optative.
ViSNU IN the Egveda. By Margaret S. Gladstone (Ph.D. Cambridge,
1928).
Copy in University Library, Cambridge.
Collection of hymns and verses in EV. addressed to Ihsnu and
Indravispu, with translation and note.s. Summary of changes brought
about by ritualism in the character of Yisnu attested in AY., YY..
SY., Brahmanas. Sutras, Brhaddevata. Further developments in
Mahabharata and Eamayana.
Some Prorlems in the nasalization of Marathi. By Bidyadhan
Nahar Sardesai (M.A., 1929).
Published in JRAS. 1930, pp. 537-565.
Accent in Gujarati. By Narhar Govind Saswadkar (M.A., 1929).
Copy in University Library, London.
An examination of the theories as to the part played by accent
in the development of Gujarati. The penultimate accent followed bv an
initial accent. Where x represents a syllable containing a short vowel
followed by two or more consonants, we have the following de\elop-
ments(l) x — — > ^ ; (2) ; (?,) x x — > ^
Sk. -w- > Ap. -v-> Guj. -//)- in certain conditions. De.scription of
accent in modern Gujarati.
TITLES AND SUMMARIES OF THESES
289
Historical and doctrinal relation between the Hinayana
AND Mahayana Schools of Buddhism. By Nalinaksha Datta
(D.Lit., 1929).
PubJished under title : Aspects of JIahayana Buddhism and its
relation to Hinayana. (Calcutta Oriental Series, No. 23.) London ;
Luzac and Co., 1930.
A STUDY OF metres IN THE OLDER Upanisads. By Purendra
Nandkrishnalal Majmudar (II.A., 1930).
Copy in University Library, London.
Analysis of the metres of the following Upanisads ; Kena, Kathaka,
I&, Mundaka, Svetasvatara. Comparison with the corresponding
metres of BY. and t-he epics. Suggested emendations of text to correct
certain metrical irregularities.
A CRITICAL edition OF THE KaPISTHALA-KaTHA-Sa.MHITA, AsTAKA I.
By Raghu Yira (Ph.D., 1930.)
Published in Kapisthala-Katha-Sariihita, a text of the Black
Yajurveda, critically edited for the first time. (Mehar Chand Laohhman
Das Sanskrit and Prakrit Series, vol. i.) Lahore : Hehar Chand
Lachhman Das, 1932.
Index to Sayana’s Bhasya on Rgveda II. By Govinda Rangacharya
Raddi (M.A., 1930).
Copy in University Library, London.
Index of all the words explained by Sayana in his bhasya on RY. ii,
together with the explanations arranged under each word.
The DEVELOPMENT OF THE Bodhis.attva DOCTRINE. By Har Dayal
(Ph.D., 1930).
Published under title : The Bodhisattva doctrine in Buddhist
Sanskrit literature. London : Kegan Paul, Trench, Trubner and Co.,
Ltd., 1932.
Ramanuja’s conception of the Deity. By Bharatan Kumarappa
(Ph.D., 1930).
Copy in University Library, London.
Life of Ramanuja. Ramanuja influenced by views of the Deity in
(1) Upanisads, (2) Bhagavadgita, (3) Paiicaratra and Puranic
literature, (4) religion of the Alvars. Ramanuja's own Anew.
The theology of the Saivagamas : Being a survey of the doctrines
of the Saiva-Siddhanta and Yirasaivism. By Shivalingayya
Channabasawayya Nandimath (Ph.D., 1930).
Copv in LTniveusity Library*. London.
voy . VITT. P.4RT I.
19
290
TITLES AND SUMMARIES OF THESES
Full account of the doctrines of Saiva-Siddhanta and Vlra^ivism.
Bibliography of the subject and lists of Sanskrit and Kanarese texts.
An Examination of Sikhavalanda. By M, D. Ratnasuriya
(Ph.D., 1931).
Copy in University Library, London.
New edition of the tenth-century Singhalese text MulsiJcka or
Sikhm-alanda, based on three MSS. in the Neville collection of the
British Museum. The introduction contains a discussion of the
interpolated passages and their origins, a note on the date, and a
systematic collection of grammatical data. Text with apparatus criticus
and a translation. Complete glossary with full references to the text,
and etymological indications.
A STUDY OF GrjAE.ATI IN THE SIXTEENTH CENTURY (V.S.) AND AN
EDITION OP Bal.^vabodha TO Upade^am.Ila. By Trirabaklal
Nandikeshwar Dave (Ph.D., 1931).
Published under title : A study of the Gujarati language in the
sixteenth century (v.s.). (James G. Forlong Fund, vol. xiv.) London :
Royal Asiatic Society, 1935.
An historical Survey of the Pa.^car.Itra religion. By
Dhirendralal De (Ph.D., 1931).
Copy in University Library. London.
Historical account of Pancaratra philosophy. Krsna-Vasudeva
an historical figure. Short account of the four principal Sariipraddyas
of Bhagavatisra : Sri, Sanakadi. Brahma, Rudra. The Bhagavata
philosophy as described in the Narayanlya episode. The philosophy
of the Pancaratras as described in their Sariihitas.
Interpretation of some doubtful words in the Atharvaveda.
By Tarapada Chowdhury (Ph.D., 1931).
Published in the Journal of the Bihar and Orissa Research Society.
vol. x\di, 1930-1, part i, pp, 25-100.
Early Buddhist Ballads and their relation to the older
Upanishadic literature. By Sumitrarao Mangeshrao Katre
(Ph.D., 1931).
Copy in University Library, London.
Nature, origin, and growth of the Buddhist ballad. Social conditions
of the period under consideration, and the literature known to the
ballads. Full treatment of some fundamental problems both in the
Upanishads and in the ballads, such as the nature of the atraan, the
TITLKS AND SUMMARIES OF THESES
291
theory of karman, and the problem of existence and emancipation.
Conclusion that the majority of the Upanishads are older than the
Buddhist ballads and that there is a marked influence of the Upanishads
on the ballads.
State of Buddhism in Ceylon as revealed by the Pali
Commentaries of the fifth century a.d. By K. K. D. E. \V. B.
Adikaram (Ph.D., 1933).
Copy in University Library, London.
The Pali commentaries, their character and sources, the various
schools of bhanakas ; a short survey of the spread of Buddhism in
Ceylon after Mahinda’s arrival. Writing down of the texts and growth
of dissentient schools. Main chapter describes the Buddhist faith in
the fifth-century as preserved in the great Buddhist centres (viharas),
marking the golden era of Buddhist learning and preaching in Ceylon.
Short historical description of the development of the ritual and the
position of the deities in the religion of that time.
The editing of the Petakopadesa. With critical apparatus and
commentary. By Arabinda Barua (Ph.D., 1933).
Copy in University Library, London.
Critical edition from four MSS. of Pefakopadesa. Introduction
discusses date and authorship. References have been traced for most
of the quotations from other texts.
Indo-Aryan loanwords in Mal.ayalam. With a study of some
Dravidian loans in Sanskrit. By Kilimanur Godavarma (Ph.D.,
1933).
Copy in University Library, London.
Extract under title : The change of a to e in the Indo-Aryan
loanwords in Malayalam. BSOS., VIII, 2.
Indo-Aryan loanwords in Malayalam classified under three heads :
Sanskrit (mostly only old loans). Middle Indo-Aryan (Prakrit), Modern
Indo-Aryan. Historical conditions of their entry into the language.
Loans from MidlA. due to Jaina and Buddhist influence. Some
MidlA. forms not otherwise recorded. Systematic discussion of the
phonetic changes in the loanwords. Index containing about 1,750
loanwords.
Studies in the word-order of Sanskrit prose. By Prabodh
Chandra Lahiri (Ph.D., 1933).
Copy in University Library, London.
Collection of references to word-order in orthodox svstems of
292
TITLES AND SUMMARIES OF THESES
grammar and works on poetics and logic, followed by statistical
examination of word-order in prose passages selected from Aitareya
Brahmana, Mahabharata, Rock and Pillar Edicts of Asoka, Majjhima-
Nikava, and Dasakumaracarita. Occasional comparisons with word-
order in IJodlA., esp. Bengali.
An edition of Nala ki bata. By Bhairaoprasad Shiwnath Pandit
(Ph.D., 1933).
Copy in University Library, London.
Extract under title : S}Titax of the past tense in Rajasthani.
BSOS., VIII, 2.
Text and translation of an East Rajasthani MS. (Roval Asiatic
Society, Todd Coll. No. 81) dating from latter half of eighteenth
century a.d. Full and systematic grammatical analysis. Vocabulary
containing all words in the text with full references thereto, and
etymological indications.
A GRAMMAR OF THE Old Kanare.se In.sc'ription.s. Including a study
of the Sanskrit and Prakrit loanwords. Bv Agaram Narasimha
Pandit Nara.simhia (Ph.D., 1933).
Extract under title : The hi.story ofp in Kanarese. BSOS.. VIII, 2.
Copies in University Library and School of Oriental Studies,
London.
Descriptive grammar of the language of the Kanarese inscriptions
of the sixth and seventh centuries a.d. Discussion of the relationship
between some of the forms and those of Modern Kanarese. Some
remarks on word-order. Discussion of the origin and .subsequent
history of Old Kan. p and pp, r. v, double consonants and consonant
groups. Text of the inscriptions studied. Inde.x of all words in the
in.scriptions, with etymological notes.
Studies in Indian dancing. As depicted in painting and .sculpture
and the representation of the musical rdgas in painting. By
Moti Chandra (Ph.D., 1934).
Copv in University^ Library', London.
Description of dancing and musical scenes in paintings and
sculpture ; description of musical instruments with their Sanskrit
and vernacular names. Connection between Rajput painting and
Vaisnava poetry in Brajabhasa. Attempt to date the Ragamala
paintings on stylistic grounds. Brajbhasa text and English translation
of all thirtv-.six raga.s and raginls in British Museum Or. 2821. Or.
8838, Or. 8839, and Add. 26550.
TITLES AND SUMMARIES OF THESES
293
An edition of the Padavali of Govindadasa. By Sudhakar Jha
(Ph.D.. 1934).
Copy in University Library, London.
Text and translation of a hitherto unput)lished Maithill ]\IS. of
1730 a.d. in the Raj Library, Darbhanga, containing a poem on Radha
and Krsna. Grammatical analysis. Historj^ of the development of
final syllables in Maithill. Sjmtax of verbal forms. Vocabulary of
aU words in the text with full references thereto and etjTnological
indications.
The position of women in the Vedic ritual. By Jatindrabimal
Chaudhuri (Ph.D., 1934).
Copy in University Library, London.
Based mainly on the original texts. Mantra. Brahmana. and
Sutra, it includes material from some less known Orhija- and ^rauta-
sutras. and from MSS. of unpublished Paddhatis and commentaries.
Exact references are given to all authorities. Separate chapters on
Daughter, Wife, and Mother.
The language of the Kharosthi Docu.ments from Chinfse
Turkestan. By T. Burrow (Ph.D. Cambridge, 1935).
Copy in University Library, Cambridge.
Detailed discussion of the phonology and grammatical forms of the
language of the Kharosthi Inscriptions discovered b>j Sir Aurel Stem
in Chinese Turkestan, ed. Rapson, Boyer, Senait, and Noble. Trans-
lation and discussion of numerous passages. Index of all words
referred to in the text.
BOOKS RECEIVED FOR REVIEW
Ahmad Kasrawi. Five hundered year's (sic) history of Khiizistan.
Persian text. Teheran ; London ; Luzac, 1934. 6s.
Akhbab AR-RApi WAL-MUTTAKi. From the Kitab al-awrak by Abu
Bakr Muhammad b. Yahya ar-SulL Arabic text edited by
J. Heyworth Dunne. London : Luzac, 1935. 10s.
Al’amuraxt duxiva DA XA -MUT.ANE. Hausa tales. Published by the
Translation Bureau, Zaria, 1935.
Beebera, D. Giuseppe M. Arabo e Berbero nel linguaggio Italo-
Siculo saggio. Beyrouth : Imprimerie catholique, 1935.
G. Bergstrasser’s Gruxdzuge des islamischex Rechts. Bearbeitet
und herausgegeben von Joseph Schacht. Lehrbiicher des Seminars
fiir orientalische Sprachen zu Berlin. Band xxxv. Berlin, 1935.
Chakrabarti, a. Cultural Fellowship in India. Calcutta : Thacker,
Spink, 1934. 6s.
Chinese and English Modern Military Dictionary. Compiled
by J. V. Davidson-Houston and Lieut. R. V. Dewar-Durie.
Peiping, 1935. 20.s.
CoLLUJi, V. C. C. The Tresse Iron-age Megalithic Monument. Oxford
University Press, 1935. 10s. 6d.
CooMARASWAMY, Anaxda K. Elements of Buddhist Iconography.
Harvard University Press, Cambridge, Massachusetts. London :
Oxford University Press, 1935. 15s.
CooMARASWAMY, Ananda K. The Rg Veda as Land-nama-bok.
London : Luzac, 1935. .3s. 6d.
Da^opanishads wtth the Commentary of Sri Upanishad-Brahma-
Yogin. Volume 1. Edited by the Pandits of the Adyar Library
under the supervision of Profe.ssor C. Kunhan Raja. Adyar,
Madras, 1935. Rs. 4.8.
Driscoll, Lucy, and Kexji Toda. Chinese Calligraphy. United
States of America : University of Chicago Press. Great Britain
and Ireland : Cambridge University Press, 1935. 9.s.
The Doctrine of the SufI.s (Kitab al-Ta‘arruf li madhhab ahl
al-tasawwuf). Translated from the Arabic by Arthur John
Arberry. Cambridge University Press, 1935. 10s. 6d.
Goodrich, Luther C. The literary inquisition of Ch‘ien-lung,
Baltimore, 1935.
BOOKS RECEIVED FOR REVIEW
295
Gopal, M. H. Mauryan public finance. London : George Allen and
Unwin, 1935. 12s. 6d.
A Grammar of the Braj Bhakha by Mirza ^an (a.d. 1676).
The Persian text critically edited . . . with introduction, trans-
lation, and notes . . . by M. Ziauddin. Calcutta : Visva-Bharati,
1935.
ELaslund, Henning. Men and gods in Mongolia. London : Kegan
Paul, 1935. 15s.
Johnson, F. Kamusi ya Kiswahili yaani Kitabu cha inaneno ya
Kiswahili. Swahili-Swahili dictionary. London : The Sheldon
Press, 1935. 2s. 6d.
The Journal of Robert Stodart. Being an account of his
experiences as a member of Sir Dodmore Cotton's Mission in
Persia in 1628-9. Published from the unique manuscript in the
Bodleian Library, with an introduction and notes by Sir E.
Denison Ross. London : Luzac, 1935. 5s.
Kalami Pir. a treatise on Ismaili doctrine. Edited in the original
Persian and translated into English by W. Ivanow. Islamic
Research Association, No. 4. Bombay, 1935. Rs. 6.8.
Katzenelenbogen, Uriah. The Daina. An anthology of Lithuanian
and Latvian Folk-songs, with a critical study and preface.
Chicago ; Lithuanian News Publishing Company, 1935.
Kemp, P. Healing Ritual. Studies in the technique and tradition of
the Southern Slavs. London : Faber and Faber, 1935. 21s.
Krishnalal M. Jhaveri. The present state of Gujarati Literature.
University of Bombay, 1934.
La Vallee Poussin, L. de. Dynasties et Histoire de I'lnde depuis
Kanishka jusqu'aux invasions musulmanes. Paris : Boccard, 1935.
Lounsbery, G. Constant. La Meditation bouddhique. Paris : Adrien-
Maisonneuve, 1935. 20 frs.
Legenden aus dek Fruhzeit des japanischen Buddhismus.
Nippon-koku-gembo-zenaku-ryo-i-ki. Ubersetzt, eingeleitet und
erlautert von Dr. Hermann Bohner. 2 parts. Tokyo ; Deutsche
Gesellschaft fiir Natur- und Volkerkunde Ostasiens, 1934.
Manichaische Handschriften der staatlichen Museen Berlin. Heraus-
gegeben im Auftrage der Preussischen Akademie der Wissen-
schaften unter Leitung von Professor Carl Schmidt. Band 1,
Kephalaia. Lieferung 1/2. Stuttgart, 1935. RM. 12.
Margoulies, Georges. Petit precis de grammaire chinoise ecrite.
Paris : Adrien-Maisonneuve, 1934. 10 frs.
296
BOOKS RECEIVED FOB REVIEW
Mees, Gitalthebus H. Dharma and Society. The Hague ; H. V.
Service ; London ; Luzac, 1935. 9s. 6d.
O’jVIalley, L. S. S. Popular Hinduism. Cambridge University Press,
1935. 7s. 6d.
Rempis, Christi-W H. ‘Omar Chajjam und seine Vierzeiler. Tubingen ;
New York ; Verlag der deutschen Chajjam-Gesellschaft, 1935.
The Rgvedabhasya of Skaxdasv.\mix. Edited by C. Kunhan Raja.
CTniversity of Madras, 19.35. Rs. 6.
Ross, Sir E. Dexisox. Marco Polo and his book. From the Proceedings
of the British Academy, Vol. XX. London : Humphrey Milford,
1935. l.s. 6(1
Selig.max, C. G. Les races de I'Afrique. Preface et traduction de
Dr. George Montandon. Paris : Payot, 1935.
Smith, IMargaret. An early mystic of Baghdad. London : The
Sheldon Press, 1935. lo.v.
SzAPSZAL. H. Ser.ua. Proby literatury ludowej turkow z azer-
bajdzanu perskiego. Krakow, 1935.
Three Tre.uises ox Mysticism by Shihabuddix SuhrawerdI
Maqtul. Edited and translated by Otto Spies and S. K. Khatak.
Stuttgart ; W. Kohlhammer, 1935. RM. 8.
Tulsidasa. Book of Ram ; the Bible of India. Rendered into English
by Hari Prasad Shastri. London ; Luzac, 1935. 2s, 6d.
Tuzak-i-W.IlajahI of Burh.\x ibx Hasax. Part I, Translated into
English by S. Muhammad Husayn Nainar. University of Madras,
1934. Rs. 5.
V.4JASAXEYI Prati.s.Ikhya OF Katyayana. 3Vith the commentaries of
Uvata and Anantabhatta. By V. Venkatarama Sharma.
University of Madras, 1934. 6s.
Wins, J. De nominale klassificatie in de afrikaansche negertalen.
Academisch Proefschrift. Nijmegen. 19.35.
Zajaczkow.ski, Axax.jasz. Studja nad jezykiem staroosmahskim.
Prace Komisji Orjcntali.stycznej, No. 17. W. Krakowie, 1934.
Kegan Paul, Trench, Trubner & Co., Ltd.
ORIENTAL AND AFRICAN BOOKSELLERS AND PUBLISHERS
CLASSIFIED STOCK OF NEW AND SECOND-
HAND BOOKS ON THE LANGUAGES AND
LITERATURES, ARTS, HISTORY, GEOGRAPHY,
AND ANTHROPOLOGY OF ASIA AND AFRICA
INSPECTION INVITED
FOR THE MOST COMPLETE RECORD OF NEW LITERATURE
on the above subjects
for kegan PAUL’S LIST OF NEW BOOKS
Catalogue XXVIII. Africa and African Languages, in the press
Libraries and collections of books and journals purchased
38 GREAT RUSSELL STREET, LONDON, W.C. I
OPPOSITE THE BRITISH MUSEUM
INDIAN ART AND LETTERS
(INDIA— CEVLON— JAVA— SIAM— IN'DO-CHINA)
(Founded I9z5)
An Illustrated Magazine published twice
yearly by the India Society (London)
The following are among recent contributors to its pages :
Dr. Charles Otto Blagden, Dr. F. D. K. Bosch, Mr. Percy
Brown, M. Georges CfEDEs, Miss Cuisenier, Prof. A. Foucher,
Mr. O. C. Gangoly, Mr. Ajit Ghose, M. Rene Grousset,
M. Joseph Hackin, Sir Kailas N. Haksar, Mr. Lionel Heath,
Miss Suzanne Karpeles, Mr. H. V. L.a.nchester, Prof. Sylvain
Levi, Mr. N. C. Mehta, Prof. Paul Pelliot, Mr. Stanley Rice,
Dr. Willem Stutterhei.m, Colonel Th. Van Erp, Mr. J. V. S.
Wilkinson, Sir John Woodroffe.
Obtainable through “THE INDIA SOCIETY”
3 Victoria Street, London, S.W., and the usual Agents
Five Shillings per issue
Ten Shillings per annum
HEFFER’S BOOKSHOP
W E carry a large stock of Oriental Litera-
ture, including Grammars, Text-books and
Translations, etc., of which Catalogues are
issued periodically and sent post free.
Authors’ MSS. considered with a view to publication.
We purchase for immediate cash settlement, libraries,
collections, and single volumes of interest. We will
pay high prices for fine items.
Our lists of “ Books Wanted ” sent free
* t
% on request.
W. HEFFER & SONS, LTD.,
BOOKSELLERS AND PUBLISHERS,
CAMBRIDGE, ENGLAND.
F O Y L E S
ORIENTAL
DEPARTMENT
Specializes in obtaining scarce Oriental
bwks for research workers and in making
up sets of scientific periodicals. The
department is under the direction of a
competent Orientalist.
A representative selection of Arabic literature always in stock.
119-125 CHARING CROSS RD., LONDON, W.C.2
Telephone : CERRARO SMO (ten lines).
SUBSCRIBE TO
ANTHROPOS
Intematioiial Review of Ethnology and Ungaktict.
Founded by the Rev. Prof. W. SCHMIDT, S.V.D.
Edited, with the co-operation of numerous missionaries and
savants, by the Rev. Dr. GEORGE HOLTKER.
The greatest scientific periodical dealing with every aspect of Anthropology.
Devoted to studies in Ethnology, Comparative Jurisprudence and Rdigion,
Linguistics, Physic^ Anthropology and Prehistory.
List of international contributors includes, on the on^ hand, missionaries and
laymen who are actually working in the Held, and, on the other, the qpccialtets
in the great Universities of the world.
Published since 1006 in annual volumes of 1,000 pages quarto, illustrated,
with plates and maps.
“ AnOtropot ** hat pubHAed Uu foQominf articUs in :
Voi,. XXX (1085), 1-2.
Wii:.BEi.M Koppebs : Die Indogermanenfrage im Lichte der historisdien
VOIkerkunde . . . . ' .
PaoFESOK Don Francisco Martinez Landero : lios Taoajkas 6 Sumos
del Patuca y VVampO (.Maps)
Generalarzt Dr. Leon Balner und Dr. Viktor Lebzelter : Zur
Biologic und Anthropologie der Kcnja in Nordost-Bomeo (ill., Karten)
Rev. Pbre Ed.hond C^aro : Le Muhaya (L’Afrique Orientale) (ill.. Carte)
Dr. C. Streit : Unbewegliche Korperzier in vorgeschichtlicher Zeit
Rev. S. Gnana Prakasar, O.M.I. : Root-words of the Dravidian group
of languages . . . . . . . •
P. Franz MCixer, S.V.D. : Beitrage zur Ethnogiaphie der Guarani-
Indianer im Ostlichen Waldgebiet von Paraguay (ill.) . .
Riv, Pere Joseph Ki.eb : Quelques notes sur lea coutumcs matrimoniales
des Mongols Ortos (Urdus) Sud (ill.. Carte) .....
Dr. Herbert Bauius : Sprachproben des Kaingang von Palmas . .
Wdlreur Schsodt : Fritz Graebner t (Portrat)
Analecta et Additamenta (215), Miscellanea (220), Bibliognphie (251),
Avis (325), Zeitschriflenschau— Revue des Revues (331).
“ Anthropos ” should be in every scientific library !
Annual Subscriptiaa, price Mk. 42. —
Of the earlier volumes (many of which are now available only in limited
numbers) the following can still be obtained; — ^Vol. I (1906), VI (1911), VII
(1912), VIII (1913), IX (1914), X-XI (1915-16), XII-XIII (1917-18), XIV-XV
(1919-20), XVI-XVII (1921-2), XVIII-XIX (1923-*). XX (1926), XXI
(1926), XXII (1927), XXIII (1928), XXIV (1929), XXV (1930), XXVI (IMl),
XXVII (1932), XXVIII (1933), XXIX (1934), XXX (1935).
Price: Pol. 1,60 Mk., Vole. VI— IX, 80 Mk., X/XI— XXIII, S2 3tk., XXIV—
XXX,42Mk.
ProeptcUutf gratia on rtqtuA.
rm
1
33
51
75
107
135
151
165
191
203
VERLAG DER INTERN. ZETTSCHRin „ANTHR0P0S“,
MODLING b/WIEN, ST. GABRIEL, AUSTRIA.
BLACKWELL’S
will be pleased to send their CATALOGUES of
ORIENTALIA
No. 333 (II). Dealing with the History, Religions, Arts, Antiquities,
Philology, and Literature of WESTERN ASIA, INDIA, and the
FAR EAST.
No. 356. Supplement to the above.
No. 377. SEMITICA, including EGYPTOLOGY, ASSYRIOLOGY, ASIA
MINOR, etc., etc.
In preparation. A Catalogue of books on the History, Geography, and
Languages of BRITISH INDIA, BURMA, CEYLON, etc.,
including the SANSKRIT portion of the Library of the late
Dr. JAMES MORISON, Librarian of the Indian Institute, Oxford.
— — GRATIS ON APPLICATION — —
BLACKWELL’S
New and Second-hand, University Booksellers
BROAD STREET - - - OXFORD
communications regarding
ADVERTISEMENTS
in the
'BULLETIN"
should be sent to —
P. C. BARTLETT’S
PUBLICITY SERVICE
27 Southampton Street
LONDON, W.C.2
THE DOCTRINE
OF THE SOFTS
(Eitab al-Ta ‘amif li-ma^hab
abl al-tasawwuf)
Translated from the Arabic of
Abu Bakr al-Ealabadhi
By A. J. ARBERRY
10s. 6d. net
The text, of which it is a
literal translation, is of great
importance, being the oldest
systematic treatise of its kind
in Arabic and a text-book of
Sufi doctrine.
CAMBRIDGE UNIVERSITY PRESS
A
N INVITATION
IIIIIIIIIIIIMIIIIIIIIinttllllllllilllMllllllllltllllllllllllllllllllllllMItlllKI
lllllllllllllllllltillllHIIIIillltJllllllllllllltlJfllllltlltlllllltlllftHUinttll
Authors and Publishers having Works containing Oriental, Foreign,
Diacritical, Phonetic, or Scientific characters ARE INVITED to
send their inquiries direct to the old established House of—
Stephen Jlmtin & Sons, Ltd.
The House that has specialized in Oriental and Scientific
productions since the days of the old East India Company.
Here are a few specimens : —
Arabic. Hindi
Ujiiii ^ ^ wiBiT
Hebrew. Chine.e.
mabnrinn msicra
c-ipiBrn Tiib c^^msri cn
Phonetics. Greek.
nqiBo naBdint^U ukuin3:k>?a KardXoym «.V Ttjy
yXwooav (Ivai ;^pi;(7i^oi oia ras
sQthb lama;nd:fea oe igk^D.mo ayopar
Burmese- Siamese. ^
^8c§0SCg8c33G6p8s030uS^CVJi56|^ WWiN Ltl
oacgrSi (o3cScoGcooc5)33c^sc(q|s filH-llwil 'f'llllu CTIMSU fl'ij
.Specimen Book of Types free on application.
If it’s “out of the ordinary” — it’s our speciality!
Interviews by the Company's Directors in London or Provinces by arrangement.
1 Fore Street, HERTFORD, Enqland
INDIAN AND IRANIAN
STUDIES
PRESENTED
TO
GEORGE ABRAHAM GRIERSON
ON HIS EIGHTY-FIFTH BIRTHDAY
7th January, 1936
ly Liar-^rj r’- iii
V
'*• 'x.
THE SCHOOL OF ORIENTAL STUDIES (UNIVERSITY OF LONDON'.
FINSBURY CIRCUS. EC. 2
1936
Sir George,
You more worthily than any Englishman of this age
have upheld the great tradition of Sir William Jones.
To-day on this eighty-fifth anniversary of your birth
your fellow-workers in the many fields, from which you
have reaped so rich a harvest and in which you were
often the first to turn the soil, unite in offering you their
grateful and admiring homage.
The long list of your publications, extending over
nearly sixty years of devoted labour, bears witness to
the boundless energy and enthusiasm and to the firmness
of spirit which, held undeviating on the path you have
chosen, has triumphed over every difficulty of circum-
stance. Neither age nor sickness has diminished that
enthusiasm nor dimmed that spirit. In your twenty-
third year, already a student of Sanskrit and appointed
a member of the Indian Civil Service, you left Ireland
for India, to enter upon that strenuous and fruitful
combination of the life of action and the life of letters
of which half a century later we saw the crowning achieve-
ment. In your seventy-seventh year you completed the’
last of the twenty volumes of the linguistic Survey of
India, and in your eighty-second year the fourth and
last part of your great dictionary of Kashmiri. You.
author in early manhood of the Bihar Peasant Life,
creator of the Linguistic Survey, compiler of so many
grammars of known and unknown languages, editor and
translator of so many Middle and Modern Indo-Aryan
texts, have more than any other contributed to our
knowledge of the innumerable languages and dialects of
India. Your work, beyond that of all others, has
stimulated in Indians themselves a just pride in their
own vernaculars and a deep and enduring interest in the
long history that lies behind them. Contributors to this
volume of studies in a field which owes so much to your
own work come from nearly every country of Europe
and from America, as well as from that dear land of
India, to whose service you gave the strength of your
youth and middle age. Yet they are but a small pro-
portion of the many who admire the greatness of your
achievement, who draw inspiration and encouragement
from your example, or who enjoy the inestimable privilege
of your friendship.
In conclusion we to whom was entrusted the grateful
task of editing this volume desire to express our thanks
to the Governing Body of the School and to Sir E.
Denison Ross, the general editor of the Bulletin, for
plaeing this part at our disposal, and to Miss Murray
Browne, whose labour in preparing it for press has
greatly lightened our own ; and to all our fellow-con-
tributors who have striven to make this volume worthy
of the great man and seholar to whom it is offered as a
birthday gift.
H. W. Bailey.
R. L. Turner.
LIST OF CONTRIBUTIONS
L. Alsdorf (Hamburg).
The \ asudevahindi. a Specimen of Archaic Jaina-Maharastrl 319
H. W. Bailey (London).
Yazdi 335
T. Grahame Bailey (London).
Does Kharl Boll mean nothing more than Rustic Speech ? 363
A. Barannikov (Leningrad).
Modern Literary Hindi ....... 373
K. Barr (Copenhagen).
Remarks on the Pahlavi Ligatures ^ and ^ . 391
E. Benveniste (Pari.s).
Sur quelques dvandvas avestiques ..... 405
Jules Bloch (Paris).
La Charrue vedique . . . . . .411
T. Burrow (Cambridge).
The Dialectical Position of the Niya Prakrit . . 419
Raymond T. Butlin (London).
On the Alphabetic Notation of Certain Phonetic Features
of Malayalam ....... 437
tjarl Charpentier (Upp.sala).
^akadhuma ......... 449
Suniti Kumar Chatterji (Calcutta).
Purana Legends and the Prakrit Tradition in New Indo-
Aryan ......... 457
Arthur Christensen (Copenhagen).
Some new AwromanI Material prepared from the collections
of Age Meyer Benedictsen ..... 467
A. Cuny (Bordeaux).
Les nasales en fin de mot en Sanskrit (et latin) . . . 477
Albert Debrunner (Bern).
Der Typus /udd- im Altindischen ..... 487
Franklin Edgerton (Yale University).
The Prakrit underlying Buddhistic Hybrid Sanskrit . 501
J. R. Firth (bondon).
Alphabets and Phonology in India and Burma
517
VI
LIST OF CONTRIBUTIONS
PAUE
Bernhard Geiger (Vieona).
Mittelpersisch vmok ,, Erbse (Linse ?) “ . . 547
Wilhelm Geiger (Municli).
Singhalesische Etymologien ...... 555
K. Goda Varma (Trivandrum).
The change of a to e in the Indo-Aryan Loan Words of
Malayalam ........ 559
Louis H. Gray (Columbia University).
Observations on Middle Indian Morphology . . 563
Olaf Hansen (Hamburg).
Sakische Etymologien . . . . . . .579
W. Henning (Berlin).
Soghdische Miszellen ....... 583
Ernst Herzfeld (Berlin).
Der Tod des Kambyses : hvamr.sjms amryata . . 589
E. H. Johnston (O.xford).
Bird-names in the Indian dialects ..... 599
Sten Konow (Oslo).
Note on the Ancient North-Western Prakrit . . . 603
J. H. Kramers (Leiden).
The Military Colonization of the Caucasus and Armenia
under the Sassanids ......
fSylvain Levi (Paris).
Mala \dhara
B. Liebich (Breslau).
Nochmals mleccha ••.....
D. L. R. Lorimer (London).
Nugae Burushaskicae
Heinrich Liiders (Berlin).
Zur Schrift und Sprache der Kharosthl-Dokumente
G. Morgenstieme (Oslo).
Iranian Elements in Khowar
A. N. Narasimhia (Mysore).
The History of p in Kanarese ....
Luigia Nitti (Paris).
Graminairiens tardifs et dialectes du prakrit
Hanns Oertel (Munich).
The E.\pres.sions for " The year con.si.sts of twelve months
and the like in 'S’edic Prose
613
619
623
627
637
657
673
681
685
LIST OF CONTRIBUTIONS vii
PAGE
B. S. Pandit (London).
SiTitax of the Past Tense in Old Rajasthani . . 695
Vittore Pisani (Rome).
Yedico yuh " se ipsiiin " . . . . . . 699
W. Printz (Halle).
Neue singhalesische Lautregel ..... 701
J. Przyluski (Paris) and C. Regamey (Paris).
Les noms de la moutarde et du sesatne . . .703
E. J. Rapson (Cambridge).
Sanskrit sd and sdh . . . . . . .709
Baburam Saksena (Allahabad).
Pali bhunaha ........ 713
Mariano Saldanha (Lisbon).
Historia de Gramatica Concani ..... 715
Hans Heinrich Schaeder (Berlin).
Ein parthischer Titel im Sogdischen . . .737
F. Otto Schrader (Kiel).
On the “ Uralian “ Element in the Dravida and the Munda
Languages ........ 751
0. Stein (Prague).
The Numerals in the Niya Inscriptions . 763
E. J. Thomas (Cambridge).
Tathagata and Tahagaya ...... 781
F. W. Thomas (Oxford).
Some words found in Central Asian Documents . . 789
R. L. Turner (London).
Sanskrit d-l-seti and Pali acchcUi in Modern Indo-Aryan . 795
Edwin H. Tuttle (Washington).
Some Dravidian Prefixes ...... 813
J. Ph. Vogel (Leiden).
Joan Josua Ketelaar of Elbing, author of the First
Hindustani Grammar . . . . . .817
J. Wackemagel (Basel).
Altindische und mittelindische Miszellen . . 823
Walther Wust (Munich).
Wortkundliche Beitriige zur arischen Kulturgeschichte und
Welt-Anschauung. II. . . . . • ■ 835
1. I. Zarubin (Leningrad).
Two Yazghulaml Texts ...... 875
HONOURS AND DECORATIONS CONFERRED ON
GEORGE ABRAHAM GRIERSON
Order of Merit ; Knight Commander of the Indian Empire ; Ph.D.
(Hon. Causa), Halle ; Litt.D. (Hon. Causa), Dublin ; Hon. LL.D.,
Cambridge ; Hon. D.Litt., Oxford ; Vagisa (Bihar and Ori.ssa) ; Fellow
of the British Academy ; Honorary Vice-Pre.sident of the Royal Asiatic
Society ; Honorary Fellow of the A.siatic Society of Bengal, Royal Dani.sh
Academy of Sciences, Bombay Branch of the Royal Asiatic Society ;
Correspondant etranger de I’lnstitut de France ; Corresponding Member
of the Konigliche Gesellschaft der Wissenschaften zu Gottingen ;
Honorary Member of the Nagarl Pracarinl Sabha (Benares), American
Oriental Society, Societe Finno-Ougrienne, I'Association PhontHique
Internationale, Bihar and Orissa Research Society, Modern Language
Association, Deutsche Morgenlandische Gesellschaft, Linguistic Society of
India, Bahglya Sahitya Parisad ; Foreign Associate Member of the
Societe Asiatique de Paris ; late Pre.sident of the Gypsy Lore Society ;
Prix Volney (Academic Fran 5 aise), 1905 ; Gold Medal (Royal Asiatic
Society), 1909 ; Campbell Memorial Medal (Royal Asiatic Society,
Bombay), 1923 ; Gold Medal (British Academy), 1928 ; Sir William Jones
Gold Medal (Asiatic Society of Bengal), 1929.
BULLETIN
OF THE
SCHOOL OF OEIENTAL STUDIES
VoL, VIII. Parts 2 axd 3.
Bibliography of the Published Writings of
Sir George A. Grierson
Compiled by Edith M. White
Abbreviations
BSOS. ; Bulletin of the School of Oriental Studies.
I A. : Indian Antiquary.
JAOS. : Journal of the American Oriental Society.
JASB. : Journal of the Asiatic Society of Bengal.
JEAS. ; Journal of the Royal Asiatic Society.
Mem. ASB. : Jlemoirs of the Asiatic Society of Bengal.
Pr. ASB. : Proceedings of the Asiatic Society of Bengal.
ZD MG. : Zeitschrift der Deutschen Morgenliindischen Gesellschaft.
PART I. BOOKS, ARTICLES, Etc.
1877
Are Kahdasa’s heroes monogamists ? (JASB., xlvi, Pt. 1, pp. 39-40.)
Notes on the Rangpur dialect. (JASB., xlvi, Pt. 1, pp. 186-226.)
1878
The song of Manik Chandra. (JASB., xlvii, Pt. 1, pp. 135-238.)
The song of Manik Chandra . . . Reprinted from the Journal of the
Asiatic Society of Bengal. Calcutta, 1878.
On the Rangpuri genitive. (Pr. ASB., 1878, p. 64.)
1879
Some further notes on Kalidasa. (JASB., xlviii, Pt. 1, pp. 32-48.)
Addenda to Further notes on Kalidasa. (Abstract.) (Pr. ASB., 1879.
pp. 107-8.)
vor. TOI. T^RT.S 2 ,\XI> a. 20
298
G. A. GRIERSON —
A further folklore parallel. (lA., viii, pp. 288-9.)
Proper names. (lA., \Tii, pp. 321-2. ix,. p. 141.)
1880
A plea for the people's tongue. (Calcutta Review, vol. Ixxi, pp. 151-168.)
1881
A handbook to the Kayathi character. Calcutta, 1881.
An introduction to the Maithili language of North Bihar, containing
a grammar, chrestomathy and vocabulary. (JASB., xlix and li,
Pt. 1, Special numbers, 1881-2.)
An American puzzle. (lA., x, pp. 89-90.)
Maithila folk-lore — Vararuchi as a guesser of acrostics. (lA., x,
pp. 366-370.)
Hindi and the Bihar dialects. (Calcutta Review, vol. Ixxiii,
pp. 363-377.)
1882
Manbodh's Haribans. Part I, Text. (JASB., li, Pt. 1, pp. 129-150.)
Completed in JASB., Uii. Pt. 1, Special number, pp. 1-75.
In self-defence. (With Addendum. On the opinions of the Biharis
themselves.) (Calcutta Review, vol. Ixxv, pp. 256-263.)
1883
Seven grammars of the dialects and subdialects of the Bihari language.
8 parts. Calcutta, 1883-7.
Folklore from Eastern Gorakhpur (N.W.P.). By Hugh Fraser.
. . . Edited by G. A. Grierson. (Notes on dialectic peculiarities,
by G. A. Grierson.) (JASB., lii, Pt. 1, pp. 1-32.)
Essays on Bihari declension and conjugation. (JASB., lii, Pt. 1,
pp. 119-159.)
1884
'/ Translation to Manbodh's Haribans. Index to Man'bodh's Haribans.
(JASB., liii, Pt. 1. Special number, pp. 1-75. Part 1 in JASB., li
Pt. 1, pp. 129-150.
Twenty-one Vaishnava hymns. Edited and translated. (JASB., liii,
Pt. 1. Special number, pp. 76-94.)
The song of Bijai Mai. Edited and translated. (JASB., liii, Pt. 1,
Special number, pp. 94-150.)
^ Some Bihari folk-songs. (JR AS.. 1884, pp. 196-246.)
Mis'kaut = Committee. (Panjab Notes and Queries, II, No. 288.)
BIBLIOGRAPHY
299
v/ Birth custom — opprobrious names — Bihar. (Panjab Notes and
Queries, I, No. 768.)
1885
Bihar peasant hfe, being a discursive catalogue of the surroundings
of the people of that province. Calcutta, 1885.
A comparative dictionary of the Bihari language. Compiled by A. F.
Rudolf Hoemle . . . and George A. Grierson. (Index to the
Ramayan of Tulsi Das.) Part I, Calcutta, 1885. Part II, Calcutta,
1889. No more published.
Selected specimens of the Bihari language. Edited and translated.
(ZDMG., xxxLx, pp. 617-673, Part I, The Maithili dialect, xliii,
pp. 468-524, Part II, the Bhoj'puri dialect.)
The battle of Kanarpi Ghat, edited and translated by ^ri Narayan
Singh and G. A. Grierson. (JASB., hv, Pt. 1, pp. 16-35.)
Two versions of the song of Gopi Chand. Edited and translated.
(JASB., hv, Pt. 1, pp. 35-55.)
Curiosities of Indian literature. (lA., xiv, pp. 124, 206, 236, 261, 292,
323. XV, 281-2, 318-19, 348. xvi, 46-8, 78, 199-200, 226-7, 256,
284, 315. xvii, 60, 88.)
Vidyapati and his contemporaries. (lA., xiv, pp. 182-196.)
The song of Alba’s marriage ; a Bhojpuri epic. (lA., xiv, pp. 209-227.)
A summary of the Alba Khand. (lA., xiv, pp. 255-260.)
1886
Some Bhoj’puri folk-songs. (JRAS., 1886, pp. 207-267.)
An introductory note by G. A. Grierson. [To “ An English-Gipsy
index. Compiled by Mrs. Grierson ”.] (lA, xv, pp. 14-16.)
Progress of European scholarship. (lA.. xv, pp. 342-5. xvd, 75-8,
110-12, 198-9, 281^, 313-15. xvii, 24-9, 321-8. xviii, 26-30,
88-90, 125-7, 246-8. xix, 72-5. xx, 116-19, 293-7, 394-6.
xxiii, 109-12.)
1887
Hints to oriental students. No. 1. Some useful Hindi books. (JRAS.,
1887, pp. 138-144.)
Gipsies in England and in India. (lA., xvi, pp. 35-Al.)
Arabic and Persian references to Gipsies. (lA., xvi, pp. 257-8.)
Language of Magahiya Dorns. (lA., xvi. p. 284.)
The mediaeval vernacular
1888
hterature
of Hindustan, with special
reference to TuTsi Das. (Verhandlungen des vii. Intemationalen
300
G. A. GRIERSOX —
Orientalisten-Congresses, Arische Section, pp. 157-210.) Wien,
1888.
The inscriptions of Pivadasi, by E. Senart. . . . Translated by G. A.
Grierson. (lA.. xvii, pp. 303-7. xviii, 1-9, 73-80, 105-8. 300-9.
xLx, 82-102. XX. 154-170, 229-266. xxi, 1-13, 85-92, 101-6.
145-155, 171-7, 203-10, 243-250, 258-276.)
Donas, Jafs, and the origin of the gypsies. Extracted from the Indian
Antiquary, 1886-87. (Journal of the Gv'psy Lore Society, i.
pp. 71-6.)
The Vikrama Samvat. (Athenaeum, 9th June, 1888, pp. 727-8.)
1889
The modern vernacular literature of Hindustan. (JASB., Ivii, Pt. 1,
Special number.)
The Danish Royal Academy's prize regarding the philological position
of Sanskrit in India. (lA., xviii, pp. 124-5.)
Gustave Garrez. [Obituary notice.] (lA.. xviii, pp. 378-9.)
The genitive in Gypsy. Extracted from The Indian Antiquary, 1887,
(Journal of the Gypsy Lore Society, i, pp. 97-9.)
Beng. (Journal of the G}'psy Lore Society, i, p. 118.)
1890
A grammar of the dialect of Chhattisgarh in the Central Provinces.
Written in Hindi by Mr. Hiralal Kavyopadhyaya . . . translated
and edited by George A. Grierson. (JASB., lix, Pt. 1, pp. 1-49,
101-153.)
Sir M. Monier- Williams on transliteration. [A letter.] (JRAS., 1890,
pp. 814-820.)
Notes on Pali and Prakrit. (Academy, 1890. 25th October, p. 369.)
1891
Joseph and his brethren. (Calcutta Review, vol. xcii, pp. .351-360.)
1892
An ancient form of submission. (.Vcademy. 1892, 2nd July, p. 15.)
1893
Notes on the district of Gaya. Calcutta, 1893.
On the early study of Indian vernaculars in Europe. (,IASB., Ixii
Pt. 1, pp. 41-,52.)
f A specimen of the Padumawati. Analysis of the Padumawati. (JASB
l.xii, Pt. 1, pp. 127-210.)
BIBLIOGRAPHY
301
Obituary notice of the death of Mr. F. S. Growse. (Pr. ASB., 1893,
pp. 119-120.)
Notes on Tul’si Das. (lA., xxii, pp. 89-98, 122-9, 197-206, 225-236,
253-274.)
1894
On a stone image of the Buddha found at Rajagrha. (JASB., Ixiii,
Pt. 1, pp. 35-8.)
Professor Weber on the Kavyamala. (lA., xxiii, p. 28.)
The Bhasha-bhushana of Jas'want Singh. Edited and translated by
G. A. Grierson. (lA., xxiii, pp. 215-222, 225-238, 265-296, 305-333,
345-352.)
The hemp plant in Sanskrit and Hindi literature. (lA., xxiii. pp. 260-2.)
Professor W. Dwight AVhitney. [Obituary notice.] (lA., xxiii,
pp. 263-4.)
1895
On the condition of Asoka inscriptions in India. (Actes du Dixieme
Congres International des Orientalistes, Deuxieme partie.
pp. 147-150.) Leide, 1895.
On the phonology of the modern Indo-Aryan vernaculars. (ZDMG.,
xlix, pp. 393-421 and 1, pp. 1-42.)
On pronominal suffixes in the Kagmiri language. (JASB., Ixiv, Pt. 1,
pp. 336-351.)
On the radical and participial tenses of the modern Indo-Aryan
languages. (JASB., Lxiv, Pt. 1, pp. 352-375.)
Note on an early supposed Bangall version of the Lord’s Prayer.
(Pr. ASB., 1895, pp. 88-91.)
[A reduced facsimile of the grant of BisapI by ^iva Sirnha to Vidyapati.]
(Pr. ASB., 1895, pp. 143-4.)
On the stress-accent in the modern Indo-Aryan vernaculars. (JR AS..
1895, pp. 139-147.)
The geographical distribution and mutual affinities of the Indo-
Aryan vernaculars. (Calcutta Review, vol. ci, pp. 258-274.)
The Tenth Congress of Orientalists, Geneva, 1894. (lA., xxiv,
pp. 136-140.)
Essays on Kasmiri grammar. By . . . Karl Friedrich Burkhard.
Translated and edited, with notes and additions, by G. A.
Grierson. (lA., xxiv. pp. 337-347. xxv, 1-9, 29-35, 85-102,
165-173, 193-9, 208-16. xxvi, 188-192. .xxvii, 179-193,
302
G. A. GRIERSON
215-221, 228-232, 309-317. xx^^ii, 6-13, 85-93, 169-179,
219-223, 247-252, 269-270. xxix, 1^.)
The Bower manuscript. (lA., xxiv, p. 370.)
Curiosities of Indian literature. Selected and translated by G. A.
Grierson. . . . Edited . . . by . . . Eamadina Sinha. [From
lA., xiv-xvii.] Bankipore, 1895.
1896
The Padumawati of Mahk Muhammad Jaisi. Edited, with a com-
mentary, translation, and critical notes, by G. A. Grierson . . .
and Mahamahopadhyaya Sudhakara Dvivedi. Vol. i, Text,
commentary and critical notes. Cantos i-xxxv (Vv, 1-286.)
(Bibliotheca Indica.) Calcutta, 1896-1911. Only six fasciculi
published.
The Satsaiya of Bihari, with a commentary entitled the Lala-candrika
by Q’ri Lallu Lai Kavi. . . . Edited with an introduction and
notes by G. A. Grierson. Calcutta, 1896.
On irregular causal verbs in the Indo-Aryan vernaculars. (JASB., Lxv,
Pt. 1, pp. 1-5.)
On the Kagmlrl vowel-system. (JASB., lxv, Pt. 1, pp. 280-305.)
A list of Ka§mlri verbs. (JASB., lxv, Pt. 1, pp. 306-389.)
Report on investigation at Bodh-Gaya. (Pr. ASB., 1896, pp. 51-61.)
Assamese literature. (lA., xxv, pp. 57-61.)
1897
The Kafmlra^abdamrta. A Ka^mM grammar written in the Sanskrit
language by I§vara-Kaula. Edited with notes and additions by
G. A. Grierson. Calcutta, 1897-8.
On the Kajmlrl consonantal system. (JASB., Ixvi, Pt. 1, pp. 180-4.)
Account of the proceedings of the International Congress of Orientalists
held at Paris. (Pr. ASB., 1897, pp. 132-142.)
1898
On the Ka 5 mirl noun. (JASB., Ixvii, Pt. 1. pp. 29-98.)
Note on a dialect of Gujarati discovered in the [Bengal] District of
Midnapur. (JASB., Ixvii, Pt. 1, pp. 185-192.)
On primary suffixes in Ka 9 mlrl. (JASB., Ixvii, Pt. 1, pp. 193-220.)
On secondary suffixes in Ka^mlrl. (JASB., Ixvii, Pt. 1, pp. 221-255.)
Note on the date of the composition of the Kavitta Ramayapa of
Tulsi Das. (Pr. ASB., 1898, pp. 113-15.)
BIBLIOGRAPHY
303
A fuither note on Tulasi-dasa and the plague in Benares. (Pr. ASB.,
1898, pp. 147-8.)
In memoriam G. Biihler. On some Swat languages. (lA., xxvii,
pp. 373-382.)
On the Kurmis of Bihar, Chutia Nagpur and Orissa. (JASB., Ixvii,
Pt. 3, pp. 110-11.)
Indian research inEussia. (Englishman, 28th February, and 1st March.
1898. )
1899
A handbook of the Kaithi character. Second, revised, edition. Calcutta.
1899.
Essays on Ka 9 mlrl grammar. [Reprinted from JASB., Ixv-lxviii.]
London and Calcutta, 1899.
On the Ka 9 mirl verb. (JASB., Ixviii, Pt. 1, pp. 1-92.)
On indeclinable particles in Ka 9 mlrl. (JASB., Ixvui, Pt. 1, pp. 93-5.)
On the genuineness of the grant of ^iva-siihha to Vidyapati-fhakkura.
(JASB., Ixviii, Pt. 1, p. 96.)
On some mediasval kings of Mithila. (lA., xxviii, pp. 57-8.)
On the Gurezi dialect of Shina. By J. Wilson. . . . Preface by G. A.
Grierson. (lA., xxviii, pp. 93-4.)
Some birth custom[s] in Bihar. (lA., xxviii, p. 195.)
On the East-Central group of Indo- Aryan vernaculars. (lA., xxviii,
pp. 262-8.)
1900
On Pashai, LaghmanI, or DehganL (ZDMG., liv, pp. 563-598.)
On the languages spoken beyond the North-Western Frontier of
India. With a map. (JRAS., 1900, pp. 501-510.)
The water of hfe. (Folk-Lore, xi, pp. 433-4.)
1901
An old Kumaimi satire. (JRAS., 1901, pp. 475-9.)
Note on the principal Rajasthani dialects. (JRAS., 1901, pp. 787-
808.)
On the history of religion in India, — a brief review. By Albrecht
Weber. (Translated from the original German by G. A. G.)
(lA., XXX, pp. 268-288.)
1902
Notes on Ahom. (ZDMG., Ivi, pp. 1-59.)
Vracada and SindhI. (JRAS., 1902, pp. 47-8.)
304
G. A. GRIERSON —
On the Mugdhavabodhamauktika, and its evidence as to Old Gujarati.
(JRAS., 1902, pp. 537-555.)
John Beames. [Obituary notice.] (JRAS., 1902, pp. 722-5.)
Note on the Kuki-Chin languages. By Sten Konow and G. A. Grierson.
(lA., xxxi, pp. 1-5.)
1903
Linguistic survey of India. 11 volumes in 20. Calcutta, 1903—1928.
The languages of India : being a reprint of the chapter on languages
contributed by George Abraham Grierson ... to the Report on
the Census of India, 1901, together with the Census statistics of
language. Calcutta, 1903.
An instance of a prosthetic g in an Indo-Aryan language. (Pr. A>SB..
1903, pp. 125-6.)
Sleeman's “ Py-khan ” — Kalidasa and the Guptas. (JRAS., 1903.
p. 363.)
Languages in India. (JRAS., 1903, pp. 425-7.)
TulasI Dasa, poet and religious reformer. (JRAS., 1903, pp. 447-466.)
A bibliography of Western Hindi, including Hindostani. [Based on
the Linguistic Survey of India.] (lA., xxxii, pp. 16-25, 59-76,
160-179, 262-5.)
The early publications of the Serampore missionaries. (A contribution
to Indian bibliography.) (lA., xxxii, pp. 241-254.)
On certain suffixes in the modern Indo-Aryan vernaculars. (Zeitschrift
fiir vergleichende Sprachforschung, xxxviii, pp. 473-491.)
1904
On the modern Indo-Aryan alphabets of North-Western India. (JRAS.,
1904, pp. 67-73.)
An Ahom cosmogony, with a translation, and a vocabulary of the
Ahom language. (JRAS., 1904, pp. 181-232.)
Guessing the number of Yibhitaka seeds. (JRAS., 1904, pp. 355-7.)
In wffiat degree was Sanskrit a spoken language ? (JRAS.. 1904.
pp. 471-481.)
Hastivanj. (JRAS., 1904, pp. 537-8.)
Linguistic relationship of the Shahbazgarhi inscription. (JRAS.,
1904, pp. 725-731.)
A few notes on the First Half of the Twenty-fifth Volume of JAOS.
Letter . . . from Dr. George A. Grierson. (JAOS., xxv, pp. 339-340.)
The languages of India and the Census of 1901. (Asiatic Quarterly
Review, xvii, pp. 267-286.)
BIBLIOGRAPHY
305
General report of the Census of India (1901). (Chapter VII, Language.)
London, 1904.
Report on the progress of the Linguistic Survey of India, presented
to the Xlllth International Congress of Orientalists. (Verhand-
lungen des XIII. . . . Kongresses, pp. 77-81.) Leiden, 1904.
“ Borrioboola-gha.” (Athenaeum, 12th November, 1904, p. 659.)
1905
Vidyapati Thakur. (JASB., New series, i, pp. 228-9.)
Pisaca = ’Dfiocftayos. (JRAS., 1905, pp. 285-8.)
1906
The Pisaca languages of North-Western India. (Asiatic Society
monographs, vol. viii.) London, 1906.
Yuan Chwang's Mo-la-p'o. (JRAS., 1906, pp. 95-105.)
The Brhaddevata and the Sanskrit epic. (JRAS., 1906, pp. 441-2.)
Dallana and Bhoja. (JRAS., 1906, pp. 692-3.)
Acjhakosikya. (JRAS, 1906, p. 693.)
The incription on the Peshawar vase. (JRAS., 1906, p. 993.)
Notes on Dr. Fleet’s article on the corporeal relics of Buddha. (JRAS.,
1906, pp. 1002-3.)
A bibliography of the Panjabi language. [Based on the Linguistic
Survey of India.] (lA., xxxv, pp. 65-72.)
Two Panjabi love songs in the dialect of the Lahnda or Western
Panjab, by Jindan. Contributed by H. A. Rose. (With some notes
by Dr. G. A. Grierson.) (lA., xxxv, pp. 333-5.)
Hinduism and early Christianity. (The East and the West, iv,
pp. 135-157.)
Do the Hindus believe in a personal God 1 [A letter.] (The East and
the West, iv, pp. 474-5.)
The languages of India and the Linguistic Survey. [Paper and dis-
cussion.] (Journal of the Society of Arts, liv, pp. 581-600.)
The languages of India and the Linguistic Survey. [Report of a
lecture.] (Indian Magazine and Review, 1906, pp. 100-2.)
1907
The imperial gazetteer of India. The Indian Empire. New edition.
Oxford, 1907-9. [Vol. i, chapter vii. Languages. Vol. ii, chapter xi.
Vernacular hterature.]
A specimen of the Khas or Naipall language. (ZDMG., Ixi, pp. 659-688.)
Vethadipa. (JRAS., 1907, p. 166.)
306
G. A. GRIERSOJ.' —
Modem Hinduism and its debt to the Nestorians. (JRAS., 190/,
pp. 311-335, 493-503.)
Eajana, Rajanya. (JRAS., 1907, p. 409.)
A verse from the Bhaktamala. (JRAS., 1907, pp. 679-681.)
The Sohgaura inscription. (JRAS., 1907, pp. 683-5.)
The Khasis and the Austric theory. (JRAS., 1907, pp. 748-750.)
An orthographic convention in the Nagarl character. (JRAS., 1907.
pp. 1057-9.)
The rain of Swati. (JRAS., 1907, p. 1060.)
Hinduism and Bhakti. [A letter.] (The East and the West, v, pp. 229-
231.)
1908
Encyclopaedia of religion and ethics. Edited by James Hastings.
Edinburgh, New York, 1908-1926.
The following articles : — Alakhnamis, Alakhgirs, or Alakhiyas :
Atheism (Indian, modern) : Atits . . . ; Baba Lalis : Bhakti-
marga : Bhils (Language) : Charan Basis : Bards ; Ganapatyas ;
Gaya : Gorakhnath : Harischandis : Kanchuliyas : Kara-
lingls: Kararls: Khakis; Khos; Literature (Indian vernacular) :
Madhvas, Madhvacharls : Maluk Basis : Nagas : Pisachas :
Prannathls ; Prapatti-marga : Radhavallabhis : Rai Basis :
Ramanandls, Ramawats : Rukhars, Sukhars, Ukhars : Sadhs :
Sadhan-panthls : Sadhu : Sakhibhavas ; Satnamis : Saura-
patas, Sauras, or Sauryas : Senapanthls : Shins : ^iva Nara-
yanls ; Tulasl-Basa.
Govinda, Gopendra, Upendra. (JRAS., 1908, p. 163.)
Vethadipa. (JRAS., 1908, p. 164.)
The modern Hindu doctrine of works. (JRAS., 1908, pp. 337-362.)
P. C. Ray’s English translation of the Mahabharata. (JRAS., 1908,
pp. 549-550.)
The nations of India at the battle between the Pandavas and Kauravas.
(JRAS., 1908, pp. 602-7.)
Robert Atkinson. [Obituary notice.] (JRAS., 1908, pp. 629-633.)
Note on Mr. Keith's note on the battle between the Pandavas and the
Kauravas. (JRAS., 1908, pp. 837-844, 1143.)
Report on the Linguistic Survey of India, presented to the Fifteenth
International Congress of Orientalists. (JRAS., 1908, pp. 1127-
1131.)
Krsna-datta Misra, Ke&va-dasa, and the Prabodha-candrodaya.
(JRAS., 1908, pp. 1136-8.)
BIBLIOGRAPHY
307
The Narayaniya and the Bhagav^atas. (lA., xxxvii, pp. 251-262,
373-386.)
India and the gypsies. (Journal of the Gypsy Lore Society, New
series, i, p. 400.)
A report ... on the Linguistic Survey of India presented ... to the
Fifteenth International Congress of Orientalists . . . 1908. (J. and
Pr. ASB., New series, iv, pp. cxiv-cxvii.)
A letter from Dr. G. A. Grierson and llr. M. L. Dames giving an
account of the [Fifteenth] International Congress [of Orientahsts].
(J. and Pr. ASB. New series, iv, p. cxvii.)
The monotheistic rehgion of ancient India and its descendant, the
modern Hindu doctrine of faith. (Abstract.) (Transactions of the
Third International Congress for the History of Rehgions, vol. ii.
pp. 44r-8.) Oxford, 1908.
The complete paper is in Asiatic Quarterly Review, xxviii, pp. 115-
Folk-etymology and its consequences. (JRAS., 1909, p. 164.)
A folk-tale parallel. (JRAS., 1909, pp. 448-9.)
Gleanings from the Bhakta-mala. (JRAS., 1909, pp. 607-644. 1910,
pp. 87-109, 269-306.)
Vasudeva of Panini IV, iii, 98. (JRAS., 1909, p. 1122. 1910, pp. 171-2.)
The Gumani Niti. Compiled by . . . Rewadhar Upreti. Communicated
by G. A. Grierson. (lA., xxx^^ii, pp. 177-188.)
Note on Professor Prince's article on the Enghsh-Rommany jargon in
vol. xxviii, 2, of the Journal of the American Oriental Society.
(JAOSt, xxix, pp. 232-5.)
Omens and folk-etymologies from Jaunsar. (Folk-Lore, xx, p. 337.)
The monotheistic religion of ancient India and its descendant, the
modern Hindu doctrine of faith. (Asiatic Quarterly Review,
xxviii, pp. 115-126.)
‘A complete report of a paper read at the Third International
Congress of the History of Religion.”
1910
The Encyclopa?dia Britannica Eleventh edition. Cambridge, 1910-
1911.
The following articles : — Ahom : Bengali : Bihari ; Gujarati and
Rajasthani ; Hindostani : Indo-Aryan Languages ; Kashmiri :
308
G. A. GKIERSON
Lalinda : Marathi : Pahari : Pisaca Languages ; Prakrit ;
Rajasthani : Sindhi and Lahnda.
An introduction to the Maithili dialect of the Bihari language as spoken
in North Bihar. ... Second edition. Parti. Grammar. (JASB..
New Series, v, Extra No. 2.)
The translation of the term “ Bhagavat ”. (JRAS., 1910. pp. 159-162.)
The modern Indo-Aryan polite imperative. (JRAS., 1910, pp. 162-3.)
The Artha-pancaka of Pillai Lokacarya. Translated by Alkondavilli
Govindacarya . . . with an introduction by G. A. Grierson. (The
Artha-pancaka of Narayana Yati. Edited by G. A. Grierson.)
(JRAS., 1910, pp. 565-607.)
Chinese riddles on ancient Indian toponymy. Note by Dr. Grierson.
(JRAS., 1910, pp. 1201-3.)
Abhinava-gupta in modern Kashmir. (JRAS., 1910, pp. 1334-8.)
A specimen of the Kumauni language. Extracts from the compositions
of Krishna Pande. Translated by Ganga Datt Upreti. Com-
municated [with an introductory note] by G. A. Grierson. (lA..
xxxix, pp. 78-82.)
1911
A manual of the Kashmiri language, comprising grammar, phrase-
book and vocabularies. 2 volumes. Oxford, 1911.
A case of Hindu s}’ncretism. (JRAS., 1911, p. 195.)
Note on Dr. Sten Konow's article onBashgali. (JRAS., 1911, pp. 195-7.)
The birthplace of Bhakti. (JRAS., 1911, pp. 800-1.)
The language of the Kambojas. (JRAS., 1911. pp. 801-2. 1912,
p. 255.)
The Takri alphabet. (JRAS., 1911, pp. 802-3.)
Foreign elements in the Hindu population. [Criticism of D. R.
Bhandarkar's paper of this title.] (lA., xl, pp. 149-152.)
1912
Etymologies tokhariennes. (Journal Asiatique, x® serie, tome xix.
pp. 339-346.)
PaisacI, Pisacas, and “ Modern Pisaca ”. (ZDMG., Ixvi. pp. 49-86.)
Pisacas in the Mahabharata. (Festschrift Wilhelm Thomsen
1912, pp. 138-141.) Leipzig, 1912.
Progress report of the Linguistic Survey of India . . . presented to
the XVIth International Congress of Orientalists . . . 1912. (J. and
Pr. ASB. New series, viii, pp. cxxiii-cxxviii.)
BIBLIOGRAPHY
309
Tengalai and Vadagalai. Note by Dr. Grierson. (JRAS., 1912,
pp. 717-18.)
Kasmiri almanacs. (JRAS., 1912, pp. 719-721.)
Progress report of the Linguistic Surv^ey of India, up to the end of the
year 1911. (JRAS.. 1912, pp. 1079-1085.)
Progress report of the Linguistic Surv'ey of India, up to the end of the
year 1911. For the XVIth International Congress of Orientalists.
(lA., xh, pp. 179-181.)
The ethnology, languages, literature and rehgions of India. Reprinted
from the Third Edition of the Imperial Gazetteer of India. Oxford,
1912, [Chapter VII, Languages, and Chapter XI, Vernacular
hterature.]
1913
Is the Ramayana of Tulasi Dasa a translation ? (JRAS., 1913, pp. 133-
141.)
On the phonetics of the Wardak vase. (JRAS., 1913, pp. 141-4.)
Alopen and SUaditya. (JRAS., 1913, p. 144.) Reprinted in lA., xlii,
p. 180.
Kanamoksa. A query. (JRAS., 1913, pp. 144-5.)
The Brhatkatha in Markandeya. (JRAS., 1913, p. 391.)
The pronunciation of Prakrit palatals. (JRAS., 1913, pp. 391-6.)
Yaska’s Datra. Shahbazgarhi and Mansehra phonetics. (JRAS.,
1913, pp. 682-3.)
Duryodhana and the Queen of Sheba. (JRAS., 1913, pp. 684-5.
1914, pp. 451-2.)
Apabhramsa according to Markandeya and “ Dhakki ” Prakrit.
(JRAS., 1913, pp. 875-883.)
1914
. . . Siva-parinayah. a poem in the Kashmiri language by Krsna
Rajanaka (Razdan). With a chaya or gloss in Sanskrit by . . .
Mukundarama Sastrl. Edited by Sir George A. Grierson.
(Bibhotheca Indica. Work No. 224.) Calcutta. 1914-1924.
Nasa = lintel. (JRAS., 1914, pp. 129-130.)
The Pahari language. [Based on the Linguistic Survey of India.]
(lA., xliii, pp. 142-151, 159-166.)
The headless horseman. (Folk-Lore, xxv, p. 382.)
1915
The North-Western group of the Indo-Aryan vernaculars. [Based
on the Linguistic Survey of India.] (lA., xliv. pp. 226-8.)
The linguistic classification of Kashmiri. (lA., xliv. pp. 257-270.)
310
G. A. GRIERSON
1916
A dictionary of the Kashmiri language. Compiled partly from materials
left by . . . Isvara Kaula, by Sir George A. Grierson . . . assisted
by . . . Mukundarama Sastri. (Bibliotheca Indica, Work No. 229.)
Calcutta, 1916-1932.
On the Sarada alphabet. (JRAS., 1916, pp. 677-708.)
1917
The home of literary Pah. (Commemorative essays presented to
Sir Ramkrishna Gopal Bhandarkar, pp. 117-123.) Poona, 1917.
The two invasion hypothesis. (JRAS., 1917, pp. 400-1.)
Mixture of Prakrits in Sanskrit plays. (JRAS., 1917, pp. 826-7.)
1918
The Ormuri or Bargista language, an account of a little-known Eranian
dialect. (Mem. ASB., vu. No. 1.) Calcutta, 1918.
Indo-Aryan vernaculars. (BSOS., vol. i, part ii (1918), pp. 47-81, and
part iii (1920), pp. 51-85.) Reprinted in lA., lx, Supplement,
pp. 6-52.
The Siva Narayanis. (JRAS., 1918, pp. 114-17.)
The Prakrit Vibhasas. (JRAS., 1918, pp. 489-517.)
1919
The Linguistic Survey of India and the Census of 1911. Calcutta, 1919.
Augustus Frederic Rudolf Hoernle. [Obituary notice.] (JRAS.,
1919, pp. 114-124.)
An Arabic word quoted by Hemacandra. (JRAS., 1919, p. 235.)
A new book in Maithili. (JRAS., 1919, pp. 235-6.)
1920
Index of language-names. (Linguistic survey of India.) Calcutta,
1920.
Ishkashmi, Zebaki, and Yazghulami. An account of three Eranian
dialects. (Prize Publication Fund, vol. v.) London, 1920.
Lalla-vakyani, or the wise sayings of Lai Ded, a mystic poetess of
ancient Kashmir. Edited with translation, notes, and a vocabulary,
by Sir George Grierson . . . and Lionel D. Barnett. (Asiatic
Society Monographs, vol. xvii.) London, 1920.
The popular literature of Northern India. (BSOS., vol. i, part iii,
pp. 87-122.)
Vocal harmony in Karen. (JRAS., 1920, pp. 347-8.)
Jompon.” (JRAS., 1920. p. 348.)
BIBLIOGRAPHY
311
On the representation of tones in oriental languages. (JRAS., 1920,
pp. 453-479.)
The home of Ramananda. (JRAS., 1920, pp. 591-6.)
Paisachi in the Prakrita-kalpataru. (lA., xlix. p. 114.)
The tale of “ The prince that didn’t exist ". (Folk-Lore, xxxi, pp. 77-81.)
1921
A grammar of the Chhattisgarhi dialect of Eastern Hindi. Originally
written in Hindi by Hira Lai Kavyopadhyaya . . . and
translated by Sir George A. Grierson. . . . Revised and enlarged by
Pandit Lochan Prasad Ka\’ya-Vinod under the supervision of
Rai Bahadur Hira Lai. Calcutta, 1921. [First published JASB.,
lis. Part 1, 1890.]
Notes on TulsI Das. [From lA., xx.] Allahabad, 1921.
Kadu and its relatives. (BSOS., ii, (1921-3), pp. 39-42.)
James Drummond Anderson. [Obituary notice.] (JRAS., 1921,
pp. 149-152.)
Sita’s parentage. (JRAS., 1921, pp. 422-4.)
Rajasekhara and the home of Paisaci. (JRAS., 1921, pp. 424-8.)
1922
Hamm- (gatau). (JRAS., 1922, pp. 77-9.)
Spontaneous nasalization in the Indo- Aryan languages. (JRAS.,
1922, pp. 381-8.)
The Apabhramsa Stabakas of Rama-sarman (Tarkavagisa). (lA., li,
pp. 13-28. lii, pp. 1-8, 187-191.)
Shahbazgarhi uthanam ; ^auraseni locative in e. (JAOS., xlii, pp. 211-
12 .)
The dragons’ teeth ; blood falling on the ground. (Folk-Lore, xxxiii,
pp. 380-1.)
1923
The lay of Alha. A saga of Rajput chivalry as sung by minstrels of
northern India. Partly translated in Enghsh ballad metre by . . .
William Waterfield. . . . With an introduction and abstracts of the
untranslated portions by Sir George Grierson. Oxford, 1923.
The lay of Brahma’s marriage ; an episode of the Alh-Khand. Trans-
lated by Sir George A. Grierson. (BSOS., ii, (1921-3), pp. 573-608.)
Hatim’s tales. Kashmiri stories and songs recorded ... by Sir Aurel
Stein . . . and edited with a translation, linguistic analysis,
vocabulary, indexes, etc., by Sir George A. Grierson. (Indian
Texts Series.) London, 1923.
312
G. A. GRIERSOX —
India and Durga in modern Hindostan. (Zeitschrift fur Indologie
und Iranistik, ii, pp. 133-9.) Part of a volume dedicated to A.
Hillebrandt.
The Eev. John Drew Bate. [Obituary notice.] (JRAS.. 1923, pp. 330-2.)
Tamil pdmbu, Sanskrit papa-. (JRAS., 1923, p. 619.)
■■ Mongoose.'" (JRAS., 1923, pp. 619-620.)
Paisachi and Chuhkapaisachika. (lA., lii, pp. 16-17.)
1924
The Prakrit Dhatv-adesas according to the western and the eastern
schools of Prakrit grammarians. (Mem. ASB., viii, No. 2.) Calcutta,
1924.
Dr. William Crooke, C.I.E. [Obituary notice.] (JRAS., 1924, pp. 147-8.)
On Sina cerebrals. (JRAS., 1924, pp. 656-663.)
On the Benares pronunciation of the Sanskrit visarga. (JRAS.,
Centenary supplement, pp. 117-121.)
1925
The eastern school of Prakrit grammarians and Paisaci Prakrit.
(Sir Asutosh Mookerjee Silver Jubilee yolumes. Vol. iii, part 2,
pp. 119-141.) Calcutta, 1925.
Prafaitica. (JRAS.. 1925, pp. 215-236.)
Once again on Sina cerebrals". (JRAS., 1925, pp. 304-14.)
On the Tirahi language. (JRAS., 1925, pp. 405-16.)
The Buddhist analogue of a Bengali story. (A supplement.) (Calcutta
Eeyiew, Third series, xiv, p. 269.)
1926
Bihar peasant life. . . . Second and reyised edition. Patna, 1926.
On the Adbhuta-Ramayana. (BSOS., iv (1926-8), pp. 11-27.)
Sortes Tulasianae. (Beitrage zur Literaturwissenschaft und Geistes-
geschichte Indiens. Festgabe Hermann Jacobi . . . 1925, pp. 449-
455.) Bonn, 1926.
The Sthiratattva inscription. (BSOS., iv (1926-8), pp. 212-13.)
Kashmiri rosf" and sosP. (JRAS., 1926, pp. 489^91.)
Kashmiri (JRAS., 1926, pp. 507-8.)
1927
The -y/acch- in modern India. (Aus Indiens Kultur. Festgabe Richard
von Garbe. . . . Veroffentlichungen des Indogermanischen Seminars
der Universitat Erlangen, Band iii, pp. 24-32.) Erlangen, 1927.
On the old North-western Prakrit. (JRAS., 1927, pp. 849-852.)
Dardic intervocalic consonants, D > L. (JRAS., 1927, p. 853.)
BIBLIOGRAPHY
313
The Sauraseni and Magadhi Stabakas of Rama-sarman (Tarkavagisa).
(lA., Ivi and Ivii, Supplement.)
1928
. . . ^ri-krsnavatara-llla, composed in Kashmiri by Dina-Natha.
Text edited, translated, and transcribed in Roman character, by
Sir George A. Grierson. (Bibhotheca Indica, Work No. 247.)
Calcutta, 1928.
The Laksmana Sambat. (JASB., New series, xxiii, p. 247.)
[Reply at the luncheon in honour of Sir George Grierson on the
completion of the Linguistic Survey of India.] (JRAS., 1928,
pp. 712-14.)
1929
The language of the Maha-naya-prakasa. An examination of Kashmiri
as written in the fifteenth century. (Mem. ASB., xi, No. 2.)
Calcutta, 1929.
Torwali, an account of a Dardic language of the Swat Kohistan.
. . . Based on materials collected in Torwal by Sir Aurel Stein.
Prize Publication Fund, vol. xi.) London, 1929.
The birth of Lorik (Magahi text). Edited by Sir George A. Grierson.
(BSOS., V (1928-1930), pp. 591-9.)
The birth of L5rik. [Translation.] (Indian studies in honour of Charles
Rockwell Lanman, pp. 243-254.) Cambridge, Mass., 1929.
Sita forlorn, a specimen of the Kashmiri Ramayana. Edited and
translated. (BSOS., v (1928-1930), pp. 285-301.)
Vararuci and the Brahmaraksasa. (Zeitschrift fiir Indologie und
Iranistik, vii, p. 96).
1930
The Kashmiri Ramayana, comprising the Sriramavataracarita and
the Lavakusayuddhacarita of Divakara Prakasa Bhatta. Edited
with an introduction and summary of the poem in EngUsh by Sir
George A. Grierson. (Bibliotheca Indica, Work No. 253.) Calcutta,
1930.
Lahnda and Lahndi. (BSOS., v (1928-1930), pp. 883-7.)
Linguistic Survey of India. Vol. i, Pt. ii. [A note.] (BSOS., v (1928-
1930), pp. 961-2.)
1931
Conjunct consonants in Dardic. (BSOS., vi (1930-2), pp. 349-368.)
Part of “ A volume of Indian studies presented ... to Edward
James Rapson . . . 1931 ”.
VOL. VIII. PARTS 2 AMD 3.
21
314
G. A. GRIERSOX —
De Lahnda, Brokpa, et quibusdam aliis. (BSOS., \d (1930-2), pp. 817-
18.)
On the modern Indo-Aryan vernaculars. (lA., Ix-lxii, Supplement.)
The tale of Bethgelert. (Folk-Lore, xlii. pp. 488-9.)
Cheques in cofiins. (Journal of the Gypsy Lore Society, Third series,
X, pp. 53-4.)
1933
Hindustan and Hindostan. (BSOS., vii (1933-5), pp. 257-260.)
Urdu AI, AU. (BSOS., \di (1933-5), pp. 260-2.)
PART 11. REVIEWS
Anderson (J. D.). A manual of the Bengali language. Cambridge,
1920. {JRAS., 1920, pp. 615-18.)
Aufrecht (Theodor). Catahgus catalogorum. An alphabetical register
of Sanskrit works and authors. Part III. Leipzig, 1903. {lA.,
xxxiii, p. 32.)
Bailey (T. Grahame). The languages of the northern Himalayas,
being studies in the grammar of twenty-six Himalayan dialects.
(Asiatic Society monographs, vol. xii.) London, 1908. {JRAS.,
1909, pp. 184-9.)
Banerji-^astrl (Anantaprasad). Evolution of Magadhi. . . . Introduction.
Oxford, 1922. {BSOS., ii, pp. 805-7.)
Barnett (L. D.). A catalogue of the Kannada, Badaga, and Kurg
books in the Library of the British Museum. London, 1910.
{JRAS., 1912, pp. 798-802.)
Barnett (L. D.). A catalogue of the Telugu books in the Library of
the British Museum. London, 1912. {JRAS., 1912, pp. 798-802.)
Barnett (L. D.) and Pope (G. Lh). A catalogue of the Tamil books in
the Library of the British Museum. London, 1909. {JRAS., 1912,
pp. 798-802.)
Bartholomae (Christian). Zum altiranischen Worterbuch. Nach-
arbeiten und Vorarbeiten. Strassburg, 1906. {lA., xxxvi, p. 180.)
Bendall (Cecil), ^'ikshasamuccaya . . . compiled by^'antideva. . . . Edited
by Cecil Bendall. (Bibliotheca Buddhica, i.) Fasciculus i. St.-
Petersbourg, 1897. {I A., xxvii, p. 84.)
Beveridge (Annette S.). The key of the hearts of beginners. . . . AVritten
down in Persian by Bibi Brooke, and translated ... by Annette S.
Beveridge. London, 1908. {JRAS., 1909, pp. 517-521.)
BIBLIOGRAPHY
315
Bhagawan Prasad (Sitarama-sarana). [The Bhakta-mala . . . with the
Tlhd of Priya-dasa, and with a commentary entitled the Bhakti-
sudha-svadu of Sri Sitarama-sarana Bhagawan Prasad (Rup-kala).
Lucknow, 1926.] {JEAS., 1928, j,. 169.)
Bhatta (Ramesvara). [The Vinaya-pattrika of Tulasi-dasa. With
a commentary by . . . Ramesvara Bhatta. Allahabad, 1913.] (BSOS.,
ii, pp. 810-12.)
Bloch (Jules). La formation de la langue marathe. Paris. 1919.
(JEAS., 1921, pp. 251-263.)
Blumhardt (J. F.). A supplementary catalogue of Hindi books in
the Library of the British Museum acquired . . . 1893-1912. London.
1913. (JEAS., 1913, pp. 1084-7.)
Bray (Denys de S.). The Brahui language. Part 1. Calcutta, 1909.
(JEAS., 1910, pp. 908-12.)
Biihler (Georg). Indian studies. No. III. On the origin of the Indian
Brahma alphabet. (Sitzungsberichte der Kais. Akademie der
Wissenschaften in Wien, philosophiscn-historische Classe. Band
cxxxii.) Wien, 1895. (lA., xxiv, pp. 246-8.)
Cowell (E. B.). The Buddha-A'arita of A-svaghosha. Edited ... by E. B.
Cowell. (Anecdota Oxoniensia.) O.xford, 1893. (I A., xxiv. pp. 179-
180.)
Crooke (W.). An introduction to the popular religion and folklore
of Northern India. Allahabad. 1894. (JEAS., 1894, pp. 874-8.)
Cummings (Thomas F.) and Bailey (T. Grahame). Panjabi manual
and grammar : a guide to the colloquial Panjabi of the Northern
Panjab. Calcutta, 1912. (JEAS., 1913, pp. 711-14.)
Darmesteter (James). Chants populaires des Afghans. Paris, 1888-
1890. (lA., XX, pp. 337-340.)
DvivedI (Sudhakara). Ramakaharu Balakanda in easy Hindi.
[Benares], 1908. (JEAS., 1909, pp. 511-16.)
DvivedI (Sudhakara) and Misra (Survaprasada.) [Manasa-pattrika.
Benares, 1904- .] (JEAS., 1905. pp. 369-370.)
Fraser (J. 0.). Handbook of the Lisu (Yawyin) language. Rangoon,
1922. (JEAS., 1923, pp. 284-8.)
Frazer (R. W.). Dravidians (South India). (Encyclopaedia of Religion
and Ethics, v, pp. 21-8.) Edinburgh, New York, 1912. (JEAS.,
1913, pp. 173-6.)
Garbe (Richard). The Saihkhya-pravacana-bhasya ... by Yijnana-
bbiksu. Edited by Richard Garbe. (Harvard Oriental Series,
vol. ii.) Cambridge, Mass., 1895. (I A., xxv, p. 232.)
316
G. A. GRIERSON—
Giornale della Societa Asiatica Italiana. Vol. iii. Eoma, 1889. (lA.,
xix, p. 286.)
Guidon (P. R. T.). TheKhasis. London, 1907. (JR AS., 1907, pp. 700-
706.)
Hoernle (A. F. Rudolf). The Uvasagadasao . . . Edited . . . with the . . .
commentary of Ahhaya Deva, and an English translation ... by
A. F. Rudolf Hoernle. (Bibhotheca Indica.) Fasciculus i.
Calcutta, 1885. (I A., xvi, pp. 78-SO.)
Hoffmann (J.). Mundari grammar. Calcutta, 1903. (JR AS., 1908,
pp. 222~S.)
Jacob (G. A.). [LaukiJca-nyaya-’njalih.] A handful of popular maxims
current in Sanskrit literature. Bombay, 1900. (JRAS., 1901,
pp. 598-9.)
Jacob (G. A.). The Vedantasara of Sadananda . . . with the commentaries
of Nrisiihhasarasvati and Ramatirtha. Edited ... by ... G. A. Jacob.
Bombay, 1894. (lA., xxiii, p. 263.)
Jacobi (Hermann). Das Ramayana. Geschichte und Inhalt . . . Bonn,
1893. (lA., xxiii, pp. 52-6.)
Kern (Hendrik). The Jataka-mala ... by Arya-^ura. Edited by Dr.
Hendrik Kern. (Harvard Oriental Series, vol. i.) Cambridge,
Mass., 1891. (lA., xxv, p. 232.)
Liebich (Bruno). Paniui. Ein Beitrag zur Kenntniss der indischen
Literatur und Grammatik. Leipzig. 1891. (lA., xxii, pp. 222-4.)
Liebich (Bruno). Sanskrit-Lesebuch. Leipzig, 1905. (lA., xxxv,
p. 184.)
Macdonell (Arthur A.). A Sanskrit-English dictionary. London, 1893.
(lA., xxii, pp. 166-3.)
Milne (Mrs. Leslie). An elementary Palaung grammar. . . . With an
introduction by C. 0. Blagden. Oxford, 1921. (JRAS., 1922,
PP- 107-9.) (BSOS., ii, pp. 339-340.)
Morgenstierne (Georg). Report on a linguistic mission to Afghanistan.
[Oslo, 1926.] (JRAS., 1927, pp. 368-375.)
[Nagarl-pracharini patrika. New series. Vol. i. Nos. 1-2. Benares,
1920.] (JRAS., 1921, pp. 286-7.)
Oldenberg (Hermann). Die Hymnen des Rigveda. Herausgegeben
von Hermann Oldenberg. Band i. Berlin, 1888. (I A., xix, pp. 286-8.)
Phillott (D. C.). Hindustani manual Third edition. Calcutta, 1918.
(JRAS., 1919, pp. 435-7.)
Pischel (R.). Grammatik der Prakrit-Sprachen. (Grundriss der
BIBLIOGRAPHY
317
Indo-Arischen PMlologie. ... 1. Band, 8. Heft.) Strassburg,
1900. (lA., XXX, pp. 553-6.)
Ratnakara (Jagannatha Dasa). [Bihari-Ratnakara, arthat Biharl-
Satsal, par Ratnakarl Tika. Lucknow, 1926.] (JRAS., 1927,
pp. 157-160.)
Rosen (Victor). Collections scientifiques de I'lnstitut des Langues
Orientales . . . iii. (Les manuscrits persans . . . decrits par . . . Victor
Rosen.) Saint-Petersbourg, 1886. (lA., xviii, pp. 95-6.)
Ryder (Arthur William). The httle clay cart. . . . Translated . . .
by Arthur Wilham Ryder. (Harvard Oriental Series, vol. ix.)
Cambridge, Mass., 1905. (JRAS., 1906, pp. 258-260.)
Sastri (Surendranath Majumdar). The dative plural in Pali. [Reprint
of an article ?] (lA., Uii, p. 68.)
Schermann (Lucian). Philosophische Hymnen aus der Rig- und
Atharva-veda-sanhita verglichen mit den Phdosophemen der
alteren Upanishad’s. Strassburg, 1887. (lA., xviii, p. 386.)
Schmidt (W.). . . . Buch des Ragawah. . . . Aus dem Mon
iibersetzt . . . von P. W. Schmidt. (Sitzungsberichte der
Kaiserlichen Akademie der Wissenschaften.) Wien, 1906. (lA.,
xxxvi, p. 180.)
Schmidt (W.). Die Mon-Khmer- Volker, ein Bindeglied zwischen
Volkern Zentralasiens und Austronesiens. Braunschweig 1906.
(JRAS., 1907, pp. 187-192.) (lA., xxxvi, p. 216.)
Sen (Dineshchandra). The Bengali Ramayanas. Calcutta, 1920.
(JRAS., 1922, pp. 135-9.)
Sen (R. R.). The triumph of Valmiki, from the Bengali of H. P.
Shastri. Chittagong, 1909. (JRAS., 1910, pp. 206-7.)
Sen Gupta (Kaviraj Dhurmo Dass). Nadi Vijnana . . . by . . .
Sankara and . . . Kanada. Translated into Enghsh from . . .
Sanskrit by Kaviraj Dhurmo Dass Sen Gupta. Calcutta, 1893.
(lA., xxiv, p. 180.)
Senart (Kmile). Notes d'epigraphie indienne, i. (Journal Asiatique,
8me serie, tome xi, pp. 504-533.) Paris, 1888. (lA., xix, pp. 43^.)
Sorensen (S.). An index to the names in the Mahabharata. Part II.
London, 1905. (lA., xxxv, p. 184.)
Temple (Sir R. C.). A native account of the thirty-seven Nats, being
a translation of a rare Burmese manuscript. (In lA., xxxv, pp. 217-
227.) (JRAS., 1907, pp. 237-242.)
Temple (Sir R. C.). The thirty-seven Nats, a phase of spirit-worship
prevailing in Burma. London, 1906. (JRAS., 1907 , pp. 237-242.)
318
BIBLIOGRAPHY
Tessitori (L. P.)- H “ Ramacaritamanasa ” e il “ Ramayana (In
Giornale della Societa Asiatica Italiana, xxiv.) (JRAS., 1912,
pj). 794— S.)
Warren (Henry Clarke). Buddhism in translations. (Harvard Oriental
Series, vol. iii.) Cambridge, Mass., 1896. (I A., xxv, p. 232.)
Weber (A.). Ahalya. ’AytAAeils, und Vervrandtes. (Sitzungsberichte
der Preussischen Akademie, 1887.) (lA., xvii, p. 302.)
Weber (A.). Die HandschriftenWerzeichnisse der Koniglichen
Bibliothek zu Berlin. Fiinfter Band. Verzeichniss der Sanskrit-
und Prakrit -HandschrLften, von A. AVeber. Zweiter Band. Berlin,
1886-1892. (lA., xvi, p. 316, xviii, p. 96, xxii, p. 112.)
AVeber (A.). Uber den Parasiprakaga des Krisbnadasa. Berlin, 1887.
(lA., xvii, pp. 273-4.)
AVeber (A.). Vedische Beitrage. (Sitzungsberichte der KonigHch
Preussischen Akademie.) Berlin, 1894. (I A., xxiv, pp. 177-9.)
AA’estcott (G. H.). Kabir and the Kahir Panth. Cawnpore, 1907.
(JRAS., 1908, pp. 245-S.)
AVhitehead (G.). Dictionary of the Car-Nicobarese language. Rangoon,
1925. (lA., Ivi, p. 214.)
AAdnfield (AA''. AA^). A grammar of the Kui language. (Bibliotheca
Indica.) Calcutta, 1928. (JRAS., 1928, pp. 931-3.)
AA'oolner (Alfred C.). Introduction to Prakrit. Lahore, 1917. (JRAS.,
1921, pp. 633-5.)
Zacharise (Theodor). . . . The Anekarthasamgraha of Hemachandra.
Edited ... by Theodor Zachariae. (Sources of Sanskrit lexico-
graphy, vol. i.) ATenna, 1893. (I A., xxiii, p. 84.)
}
The Vasudevahindi, a Specimen of Archaic Jaina-
Maharastri
By L. Alsdorf
Abbreviations used: Vh = Vasudevahindi ; M = Mahara?trl ; tTM = Jaina-
Mahara§trl ; S = Sauraseni ; JS = Jaina-SaurasenI ; AMg — Ardhamagadhi ;
P = Pischel’s Prakrit Grammar ; JErz = Jacobi’s Ausgewdklte Erzdhlungen
in Mdhdrdstrl ; Mah. Xis. = Schubring, Das Mdhdnisiha^Sutta ; Pkt, Skt
= Prakrit, Sanskrit.
TN the Introduction to his edition of the Avasyaka tales ^ the late
-*■ Professor Leumann wrote on p. 1 : “ Since — as may be easily
understood — Jacobi, when he endeavoured to utilize the language
and contents of the Jaina tales for Indology, started with the
Uttaradhyayana Tika of Devendra which was in his possession, and
since this author, belonging as he does to the period of decadence and
therefore more familiar with Sanskrit, writes a rather doubtful
Prakrit — therefore the editor was chiefly concerned with the
establishing of a thoroughly reliable Prakrit text . . . that was to be
suitable for clearing the way for a more correct judgment and
utilization of medieval Jaina Prakrit.” The appearance of the
first portion of Leumann’s text was welcomed by Pischel in his Pkt
grammar (§ 21) in the following terms : “ The most important text
in JM is : Die Avasyaka-Erzahlungen. Herausgegeben von Ernst
Leumann. 1. Heft. Leipzig 1897. The absence of any commentary
unfortunately renders the understanding rather difflcult ; some
passages remain wholly obscure. But even these few forms show that
from texts in Jll we may yet expect much new and important
material.”
In spite of all that has been done since then in the domain of Pkt
literature and grammar, the hope expressed in Pischel's last sentence
can hardly be said to have been fulfilled. The work so ably begun by
Leumann has never been continued ; the first portion of his Avasyaka-
Erzahlungen has remained the last. Many other texts have become
accessible, but they are almost without exception open to Leumann’s
criticism of Devendra : works like those of Haribhadra, Somaprabha,
etc., bear witness to the skilful handling of a fixed — not to say
“ dead ” — literary language by later authors, but they are not likely
^ Cf. the following quotation from P.
320
L. ALSDORF —
to throw much light on those problems that interest us most, i.e.
the origin and development of JAI and its relations to the other Pkt
dialects. Even those JM texts that are available have not received
systematical grammatical treatment, so that Pischel’s and Jacobi's
descriptions of the language (P and JErz), much antiquated as they
must necessarily be, are still the standard works to refer to. And there
is an almost entire lack of really old texts.
One such text has lately come to light. Five years ago there
as vols. 80 — 1 of the Atmanand Jam Granth Patnamala
the Vasvdevakindi by Sanghadasaganin.^ The date of the work is
not known, but as it is mentioned three times by the Avasyaka Curni,
‘‘ it cannot be later than the sixth century a.d., if we allow an interval
of no more than a century between it and the curnis, and as much
between them and Haribhadra. ” ^ There is, however, nothing to
prevent us from assigning to the Vh a much earlier date than the
sixth century. That it must actually be centuries older will, it is hoped,
be clear from the sketch of its language to be given below.
The Vh might well be called unique for more than one reason. The
fact that it presents us with a detailed and circumstantial Jain version
of Gunajhya’s Brhatkatha, quite independent of the Kashmirian and
Nepalese versions and highly valuable for the reconstruction of the
lost original,^ lends it the greatest possible literary importance.
Further, the tales inserted into the old commentaries are mostly written
without literary ambitions. The aim of the commentators — Devendra
being a pleasant exception — seems to be to give the facts of the stories
in the shortest possible form — the result often being a kind of telegram
style which is sometimes almost unreadable and difficult to understand.
The \Ti constitutes a continuous prose work of 370 quarto pages —
its very scope and plan has no parallel in the older non-canonical Jaina
* Edited by Munimabarajas Caturavijaya and Punyavijaya. A third fasciculus
which wa.s to contain the introduction has not appeared. The text is provided witli
vety useful indices of proper names, a pratika-list, list of inserted tales, etc. Though
i)ased on twelve M.SS. (of which readings are given) and outwardly a model performance
testifying to the praiseworthy application and zeal of the editors, the edition i.s
thoroughly uncritical and — particularly as regards the orthography — in no way better
than a moderately good il.S. The absence of any kind of commentary or explanative
glosses is all the more regrettable liccause the text is not always easy and often full
of mistakes and corruptions.
- Jacobi, Introduction to the 2nd edition of Hemacandra’s Parisistaparvan {Bibl.
Indica), p. vii. That the Av. Curni mentions the Vh not once but thrice I found in the
papers left by the late Profe.s.sor Leuraann.
^ Cf. the paper read by me at the 19th International Oriental Conference at Rome,
entitled “ Eine neue Version der v'erlorenen Brhatkatha des Gunadhya ”.
A SPECIMEN OF ARCHAIC JAINA-MAHARASTRI
321
literature. The style is far from being concise or dry ; it presents us
with a vivid, characteristic, and highly interesting picture of a hving
language. The tale is frequently embellished with flowery descriptions
so dear to Indian poets, and some of them are — at least partially —
in the old Vedha metre, hitherto unknown to occur outside the canon —
an unmistakable sign of great antiquity.
But it is the language of the Vh with which we are here concerned
and which has also quite a number of surprises in store for us. An
exhaustive and systematical description and treatment of it would hr
far exceed the space available for this paper and must be left to a future
occasion. Here I can only give, as a kind of preliminary notice, and
without any claim to completeness, a brief enumeration and discussion
of some of its more important and striking features and peculiarities.
It is the verbal system which offers the greatest variety of new
and interesting forms.
In a number of cases ^ the 1st person sing, of the present indicative
ends in -am instead of in -ami. This use of the secondary ending is
normal enough in the future (e.g. bhavissam) ; in the present tense
it is unknown to P. Yet at least one such form occurs in the canon.
In the sixth dasa of the Ayaradasao we read (cf. Schubring, Die Lekre
der Jainas, p. 181) that a layman, when asked a question, must
truthfully and openly say whether he can answer it or not : kappanti
duve hhdsdo bhdsittae, jaka : jdnam vd “jdnam’', ajdnam vd “no
jdnam ”, i.e.jdnan vd “ jdndmi ”, “ either, if he knows, ‘ I know,’ ” etc.
In Vh I have noted the following fifteen cases : jdnam 174, 26 ® ;
na jdrmm, 145, 23 ; 174, 8 ; 353, 27 ; na-ydnam, 19, 3 ® ; 83, 22 ;
115, 26; 144, 24 (v. 1. na-ydndmi)-, iccham, 4, 3; 17, 17 ; 229, 22;
350, 28 ; tattha y'aharn pdsam, 283, 16 ; tkavijjam,* 109, 7 ; jivatn,
' A general remark may not be out of place. An abnormal or somewhat strange
form need not be suspected merely because it occurs very rarely, perhaps only twice
or thrice. In the papers left by Leumann I found a slip where he had collected variants
of the Vise§avasyakabha§ya. Again and again the variant consisted in an aorist form
being replaced by the corresponding form of the present indicative. This is only one
instance of how the more uncommon forms were gradually eliminated by the scribes —
we have to content ourselves with what little the^' have left.
^ The figures refer to the pages and lines of the printed text.
® Only one MS. reads “ na-ydnarri kumdram panattkam ” ** I did not know that the
prince had disappeared.” The others have the corrected ” reading, “ tm ya
which, however, is incompatible with the following accusative “ kumdram
panattham In this and five other cases (19, 3 ; 83, 12 ; 109. 7 ; 115, 26 ; 144, 24)
a vx)man is speaking, which excludes the possibility of the form in -am being regarded
as a nom. sing, of the present participle.
^ 1st sing. ind. pass., possibly to be corrected to a 1st sing. opt. pass. *thav{jjijjam.
322
L. ALSDORF
91, 14. To these must be added two optatives : pdsijjarn, 6, 18, and
passejjain, 125, 3. A 1st sing. opt. in -jjam is also wanting in P where
we find only -jja, -jja, -jjdmi. That both forms (ind. -am, opt. -ijjam)
are genuine archaisms is proved bevond doubt by the fact that they
have counterparts in Pali (e.g. gaccham, labheyyam, cf. Geiger, Pali,
§§ 122, 127, 128). There -am instead of -ami is peculiar to the language
of the gathas, i.e. the oldest stratum of the language. The only other
Pkt text where a 1st sing. ind. in -am occurs seems to be the
Mahanisiha Sutta, from which Schubring (Mah. Xis., p. 90) quotes the
sloka-pada “ gaccham cettham surani uttham dhdvam ndsam paldmi
um ”, and three single forms, also from verses, caram, na ninhavarn,
sakkanam ( = saknmni 1). This is interesting because we shall presently
see that another peculiarity of the Mahanisiha is also shared by Vh.
In the Mahanisiha, “ the 1st plur. is often constructed with aham
or is otherwise used as a 1st sing.’’ e.g. ahayam . . . anucittlmno, ndham
. . . cukkimo, etc. In 'VTi we read, e.g. 84, 7, aham . . . dacchdnw ; 290,
28, tao ’ham tdo lavdmo ; 291, 24, lavai ya me : “ icchdmo 172,
11, samcaramdni . . . siminio; 178, 22, mayd bhaniyd : jdnlhdmo ”
tti. tao niggayd “ jdnihisi” tli vottunam.
According to P, § 457, the 1st person plur. of the atmanepada is
wanting in Pkt. In Vh I have noted it seven times, but in all cases
it is used as a 1st sing., twice even with ahain : 147, 18, aham . . .
padicchdmahe ; 206, 18, aham . . . anuvattdnmhe ; 144, 7, iitthio mi
“ kattha marine vattdmahe ? ” tli cintayanlo ; 352, 22, cintemi : “ kammi
paesammi vattdmahe ? ” ; 139, 24, ramdmahe (v. 1., °mi ya !) ; 155, 16,
uvabhunjdmahe . . bhutta-bhoyano ya . . . : 330, 19 (in a gaha !),
bhanai : “ . . . bandhdmnhe . . .
This form in -dmahe, used as a 1st sing., probably helps to explain
another very strange and hitherto utterly unknown form. Eleven
times there occurs a 1st sing, in -ahe ; acchahe, 180, 14 ; 206, 12 ;
247, 1 ; aticchahe, 319, 24 ; dsahe, 289, 13, 29 ; dissahe (pass.), 199, 6 ;
passahe, 218, 10 ; pdsahe, 293, 5 ; vattahe, 247, 12. Failing any other
explanation, I believe that -ahe is formed from -dmahe after the analogy
of the proportion : 1st sing. act. -dmi : 1st sing. med. -e (vanddmi :
rande = vanddrnahe : vandahe). And, lastly, a single form may be
noted which — if not a mere mistake — looks like a compromise between
-ahe and -e : 156, 17, 1st sing, acchae instead of acche.
According to P, § 455, the 1st plur. ind. (which is used as 1st
plur. imp. as well, § 470) ends in -mo which may be replaced by -mu
in verses only. Yet in the prose of Vh -mu occurs not infrequently,
A SPECIMEN OF ARCHAIC JAINA-MAHARASTRI
323
e.g. ind. ainemu, 108, 11; na-ydndmu, 117, 8; acchdmu, 115, 24;
karemu, 117, 17 ; imp. vasdmu, 82, 3 ; hardmu, 100, 2 ; anuvaydmu, 138,
2 ; anumaggdmii, 138, 12 ; karemu, 85, 15 ; 109, 12 ; 153, 15, etc.
Now since, as we have seen, the 1st plur. may be used as a 1st sing,
as well (probably originally as a pluralis majestatis), I have little doubt
that the 1st sing. imp. in -mu taught by the grammarians but —
according to P, § 467 — not found in literature, is nothing but a 1st
plur. used as a 1st sing. In support of this explanation I can quote
from \Ti at least one 1st sing. imp. in -mg : 122, 5, “ na me sobhai iharn
acchium, avakkamdrno ” tti, “ it is no good for me to stay here, I will
run away ! ” — Pischel believed that -mu, -su, -u as terminations of the
imperative corresponded to the indicative terminations -mi, -si, -i,
and mainly for this reason he disputed the usual derivation of -su
from Skt. -sva. If my explanation of the 1st sing. imp. in -mu is
accepted, Pischebs view of -su — not very convincing considering Pali
-ssu (cf. Geiger, Pali, § 126 and note 1) — becomes wholly untenable.*
It has till now been taken for granted that forms of the aorist
have survived in AJIg only. From Mi we learn that they occur
in archaic JM as well.* We even find in Mi several forms which are
either quite new or hitherto kno\vn only from grammarians.
The 1st sing. act. of the aorist is “ very rare and ends in -issarn ”
(P, § 516 ; e.g. akarissarn), “ with double s as in Pali (where, however,
the more usual form has a single s, cf. Geiger, Pdli, § 159, iv). One such
form occurs Vh 225, 17 : “ ndharn tuhbham kuppissam,” “ I was not
angry with you.” The context excludes every possibility of the form
being taken as a future. Yet the formal identity of a 1st sing. aor.
in -issam with the 1st sing. fut. cannot be overlooked, and I believe
that the doubling of the s (for which neither Pischel nor Geiger offer
^ W. Schubring has now traced it in the Mahanisiha, where a 1st sing. imp. demu
occurs twice {Mah. Nis., p. 91).
^ In Vh there occurs in a number of cases a 2nd sing. imp. in -asw (e.g. vaccdsu,
93, 15 ; 179, 18 ; ghaddsu 94, 24 ; kardsu 96, 10). The long d of these forms may,
of course, be explained by the influence of the parallel form in -dhi. We may, however,
also have before us the regular compensative lengthening due to the double s (which
is seen in Pali -S9u) having been simplified.
® It might be argued that such forms are to be regarded as mere
“ Ardhamagadhisms ”, which do not prove anything for JVI, This argument would
perhaps hold good if we had to do with legendary, dogmatical, or disciplinarian texts.
Where the contents are so closely akin to those of the canonical scriptures it is not
surprising if the language betrays a strong AMg influence (cf. Mah. Nis., p. 86).
But, except for some inserted legendary tales, the contents of the Vh are as secular
and non-canonical as possible, and there can be no doubt that its language is JM
throughout.
324
L. ALSDOBF
any explanation) is actually due to the influence of the similar future
form. This assumption receives a welcome support by another form
which is hitherto quite unknown and without a parallel in either Pkt
or Pali. \Ti 289, 27, we read : “ dacchiham c'aham,” " and I beheld.”
Now it is well known that besides the normal future in -issdmi {-issarn).
-issai, -issdnio, etc., there exists in Pkt a second future in -ihdmi
(-iham), -ikii, -ihdmo, etc. But if the h of these forms — as must
obviously be the case — goes back to sy > ss> s, we should expect before
it 7 (with compensative lengthening) rather than i. This form with a
long i, not recorded by the grammarians and not found in P or JErz,
actually occurs in Vh, e.g. 51, 22, ghatfiham ; 22, 28, bhiinpham ;
78, 22, jdnihdmo ; 89, 21, jmccJnhdtm ; 91, 8, jtvVid-tno ; 138, 7.
dacchihdmo, etc.^ The 1st sing, belonging to the last of these forms
would be dacchiham, i.e. the very same form we have just noted as a
1st sing. aor. This use of a 1st sing. fut. dacchiham as a 1st sing. aor.
seems to prove that a connection was felt to exist between the 1st sing,
fut. and aor. in -issam : the identity of these two forms seems to have
justified the use of any 1st sing. fut. as 1st sing. aor.
There are, however, indications to show that the relations between
the future and the aorist were not limited to the 1st person of the sing-
The grammarians know two aorists formed with h instead of s which —
to conclude from P, § 516 — have not yet been found to occur in texts,
viz. kdhi besides kdsi (from kr) and thdhi besides thdsi (from sthd).
Here the h may actually go back to the s of kdsi and thdsi. But it is
very significant that these /<-aorists, too, are identical with forms of the
future : kdhi and thdhi are perfectly normal 3rd persons sing, of the
futures kdham (P, § 533, where kdhi is actually quoted) and thdham
(P, § 524, “ thdhii ”). Two other aorist forms (vocchiya and gacchiya)
which can only be explained by referring to the futures voccham and
gaccham will be discussed below'. And, lastly, the Mahanisiha furnishes
what might be called a counter-proof. It uses [Mah. Nis., p. 91)
a 3rd plur. “ hhavisum, once even spelt bhammsxim, as if the forms were
aorists ; but the context, where also bhavihenti precedes them, proves
beyond doubt that they are futures.” Here forms of the aorist have
penetrated into the future : the exact contrary of the relations between
aorist and future w’hich we have noticed so far.
One of the most common aorist forms in AMg is vaydsi “ he spoke ”,
which is — like dsi — used for other persons, particularly the 3rd plur.,
* The “ missing link ” between -issam and -iham is supplied by the Mahanisiha,
where futures in •isarn occur, e.g. riimiccisatri sujjhisarri {Mah. iVi'.?., p. 91).
A SPECIMEN OF ARCHAIC JAINA-MAHARASTEI
325
as well. In \Ti it occurs as 3rd sing. 351, 28 ; as 3rd plur. 33, 17. But
besides it there occurs three times a hitherto unknown by-form with
samprasdrana of the root : 284, 9 and 13, 3rd sing. “ inam uddsl
(284, 9, one MS. reads “ idam vaddsi ”) ; 324, 3, 3rd plur. “ irnam
vayanam uddsi”.
Two, as it were, “ normal ” forms are the 3rd sing, velavesi (291,
20 ; from velavai, “ to upbraid ”) and the 3rd plur. vinnamnm (216,
2 ; from vinnavai vijnapayati).
A more doubtful case is found, 29, 11, where a husband, referring
to his previous warnings, says to his wife : “ kirn iddnirn rodasi ?
mamam tadd na sunesi bhannamdni ! ” “ MTiy do you weep now?
At that time you did not listen to me when I spoke to you ! ” It is,
of course, quite possible to take sunesi as a 2nd sing, of the present
indicative. But a 2nd sing. aor. (which does not differ from the 3rd)
would have the same form (cf. P, § 516, kahesi, etc. ; the final t may
also be short, cf. e.g. akdsi, ahesi, etc.), and the context seems to demand
most categorically a form of a past tense.
§ 466, end, Pischel deals with some rather mysterious forms in -la
which are “ mentioned by the grammarians as being used in the sense
of the imperfect, aorist, and perfect tenses ”, and he thinks that.
■■ inexplicable as it seems,” acchla, genhla, daliddala, maria, hasla,
huvla, dehla are really optatives of the present and kdhla, thdhla, hohla
are really optatives of the aorist. Whatever the correct explanation
of these forms may be, the Vh furnishes the first instances of their
actual use. We read there ; 289, 17, gacchlya, “ I went ” ; 289, 28,
dine gameslya, “ I spent the days ” ; 278, 32, vocchlya, “ he spoke ” ;
111, 22, kdslya, “ he made ”d According to Pischel, we should have to
regard the first of these forms as an optative of the present and the
other three as optatives of the aorist. But an aorist *vocckl can hardly
be explained except as having been derived from the future voccharn
(P, § 529), of which the 3rd sing, is vocckii, vocchl ; and this makes it
at least very probable that gacchlya does not belong to the present
gacchai, but that an aorist *gacchl had been derived firom the future
gaccharn (P, § 523). Here, again, the very close relations become
visible that must have been established in Pkt between the future
^ As in the text of Vh ca, ya, ti are frequently inserted where they are superfluous
and even manifestly impossible, there is just the possibility of declaring the final
ya of the forms in question (or some of them) to be such a spurious ya. Considering,
however, the unanimous testimony of the grammarians, I feel confident that we have
actually to read gacchiyay etc.
326
L. ALSDOEF —
and the aorist. Unfortunately our material is as yet too scanty to
allow of a more definite description of them.
The verbum substantivum calls for a few remarks. The first
point to attract our attention is the quite unusual frequency of its
use. E.g. the 2nd plur. ttha is qualified by Pischel (§ 498) as
“ very rare ”. He gives one single reference (to Setubandha 3, 3)
and in his paradigm ascribes ttha to M only. In Vh it occurs more than
forty times ! (e.g. 86, 6, 8, 16 ; 92, 21, 23 ; 93, 6 ; 96. 2, 4 ; 102,
17 ; 103, 10 ; 107, 21 ; 110, 5, 7, 24, etc.). The 1st and 2nd sing, and
the 1st plur. are even more commonly used. As we have doubtless
to do with old forms that are gradually becoming obsolete in Pkt,
their strong vitality in \Ti must be looked upon as a sign of antiquity.
Another very striking peculiarity points in the same direction. We
should expect enclitic forms like mi, si, mo, ttha to be normally placed
after the word (mostly a past part.) to which they belong, ^ and this
is actually often enough the case, e.g. gao mi. 182, 20 ; patto si, 146, 16 ;
patta mo, 148, 15 ; jdya ttha, 86, 16 ; etc. But perhaps even more
frequently the forms in question are placed before the word they belong
to, and they may even be separated from it by one or several other
words, e.g. 281, 16, tao mi niggao ; 196, 2, tattha ya mi gao ; 283, 16,
tao mi junn' anteuram gayd ; 229, 25, jai si saho puraccarane ; 80, 18.
sumarasi, jam si bdlabhdve do vi (?) Aimuttaena nabJiacdrind bhapiyii
...?■, 214, 19, tattha ya mo Mandara-samlve vutthdo ; 86, 8.
kao ttha-m-dgayd ? ; 96, 2. jam ttha devle dnattd ; 238, 4, jai ttha
jamma-marana-bdhulam sagisdram chindiu-ka.xak ; 125, 2, tumhe ttha
/nayd samgdmao padiniyattd akkhaya-sarlrd dittha, etc. ; cf. also
several of the pa.ssages quoted below, p. 329. note, as instances of
the conditional.
All these sentences would not only retain exactly the same meaning
but they would even look more normal if mi, si, mo, ttha were replaced
by the personal pronouns aham, tmmm, amhe, tumhe respectivelv.
It is thus easy to imagine that these forms of -fas could have come
to be regarded as equivalents of the nominatives of the personal
pronouns. That this has, at lea.st to some extent, actuallv been the
case is proved by four forms which the grammarians enumerate as
equivalents of aham. Pischel has pointed out (§ 417) that amhi, ammi.
mmi are = Skt. asmi, while ahammi is = aharn mi = aham asmi,
and that there is no reason to doubt the statement of the grammarians
^ Cf. the quotations in P. § 145.
A SPECIMEN OF ARCHAIC JAINA-MAHAEASTRI
327
that these forms were, used in the sense of aham ^ — even though none
of them was known to occur in a Pkt text. In Vh we read : 217, 19,
a7n}ii pesiyd siimaram-dme (sc. deme) tuhbham paya-samhmn ; 146, 2,
tao tena afnhi hhanio ; 182, 22, ten'amhi ^ nmhuram dbhattho ; 210, 24,
tena y'amhi bhanio ; 212, 7, tly' ® anmii tmhurani ahattho ; 218, 5,
pavittho y'anmii ; 230, 16, tehi y'ammi tuttheliirn anihio ; 279, 3,
tags' atmni phalam patto. It will be seen at once that these sentences
cannot be separated from those just quoted above : the use of amhi
and atmni corresponds exactly to that of mi, si, mo, ttha. In aU cases
it would be possible, but is by no means necessary, to interpret the
forms of ^/as as personal pronouns. Two less equivocal passages
seem to be 165, 18 : keriso si kesu bhavesn asi ? — kldrsas tvam kesu
bhavesv dsih ? and 217, 29 ; tao mi iittinno vlsamdmi = tato’ ham ultimo
visratndmi.*^ But 24, 3, we read: litmatn si me bhdyd kanittho dsl,
and here it seems almost unavoidable to regard si as an interpolation.
Yet even this passage may be correct ; " tumamsi ” = tvam would be
an exact counterpart of “ ahammi” taught by the grammarians as
nom. sing. = aham.
From the 1st sing, bemi = bravlmi, AMg and JM have derived
a 3rd plur. benti (P, § 494 ; Vh, 118, 11 ; 223, 15). A 3rd sing, bei,
not recorded by Pischel, occurs Vh 35, 18.
Hemacandra admits (i, 46) datla besides dinna, but according to
P, § 566, it is found only in a Pallava Grant (6, 21, data) and in proper
names. In 'STi datta and dinna are about equally frequent. In one-half
of the text I have counted the former about forty times (e.g. 106,
24, 26, 27 ; 181, 3, 5, 6 ; 241, 6, 8, etc.).
Even in Skt manye sometimes “ has almost got the character of
a particle ” (Speijer, Sanskrit Syntax, § 500, note). In Vh marine is
regularly used as such after interrogatives, where it might be translated
bv German " wohl ”. E.xamples could be quoted by the dozen, but a
few must suffice : kirn maririe, 13, 20 ; ko m., 18, 27 ; kd m., 101, 8 ;
keria m. kdranena, 133, 15, 28 ; kassa m. rdino, 83, 28 ; klsa ni., 14, 27 ;
kattha m., 20, 10 ; kiha m., 310, 15 ; kayd m., 176, 26 ; kayaro m. esa
^ The peculiar use of atthi illustrated in § 417 is also familiar to Vh : 57, 10, attki
koi . . . pariiasai ; 10, 29, n'atthi koi vahi-doso dlsai ; 125, 5, atthi me puno rajja-sirl
hnjjd ?
2 There can hardly be any doubt that we have to ^vrite ten' amhi, y'amhi, y'ammi,
etc., and not tena 'mhi, ya mhi, ya mmi,
3 tlya — tie, cf. below, p. 328 f.
* It is grammatically possible, but otherwise very unlikely that we have to do
with two sentences : tato 'smy uttirnah. visramdmi.
328
L. ALSDORF
devo, 78, 13. This use of manne throws new light on the adverb vane
taught by Hemacandra (ii, 206 ; cf. P, § 457). It seems now certain
that vaw must be explained as mane = manye.
Turning to the declension of nouns, we shall first of aU make the
important statement that the nom sing. masc. of A-stems ends in -o,
without a single exception. The nominative in -e occurs only in two
short AMg quotations — a Vaitallya stanza, p. 30, 2-3, and half a
sloka (not recognized as such by the editors), p. 234, 24 — the only ones
that I have been able to discover in the whole text. The other important
characteristic of AMg, the locative in -msi, does occur in \Ti, but of the
very few forms the majority are obvious Ardhamagadhisms, so that
the same may safely be assumed of the rest.^
In normal AMg and JM the dative sing. m. n. and the inst. gen.
loc. fem. of the A-stems end in -de, the inst. gen. loc. of feminine I-
and U-stems end in -ie, -ue. According to Pischel (§ 361) a dat. sing,
in -dya is limited to AMg verses and Magadhi verses. As to the inst.
gen. loc. sing, of the feminines, he admits -da only for M (§ 374 f.),
-la and -ua only in verses where the metre demands a short vowel
(§ 385). He evidently rejects, deeming unnecessary even to mention it,
the opinion of Leumann who maintains {Avisyaka-Erzdhlungen,
p. 3 f.) that the feminine forms in -dya, -lya, -uya, which are not
infrequently found in older texts such as the curnis and even the
bhasyas, are genuine, and that -ya is the “ older form ” (as compared
with -e) “ which Pkt has in common with Pah ”.
In Vh the following state of things prevails. Besides the dative
in -de (e.g. vahde, 326, 7 ; attJide anatthde ya, 124, 14 ; piittattde, 76,
13 ; 91, 21, etc.), that in -dya is also found : vahdya (= vadhdya),
169, 19 ; 245, 2, 4 ; 313, 5, 7 ; hiydya, 268, 5 ; uvagdrdya 163, 4 ;
vindsdya, 313, 8. Feminine forms in -dya, -lya, though very much rarer
than the normal ones in -de, -le, are also not infrequently met with
(e.g. instr. padihdrdya panaydya, 213, 13 ; gen. asuhdya, 230, 6 ;
kanndya, 311, 1 ; inst. tutthlya, 121, 31 ; buddhla, 10, 24 ; loc. veldya,
150, 20, parinaydya santiya, 173, 1, etc.). But in addition to them Vh has
preserved in a few cases a form w'hich seems to me to prove conclusively
that Leumann was right in comparing a Pkt. kanndya to Pali kanndya.
According to Geiger's grammar (§§ 81, 86), Pali forms the inst. gen.
* I have only noted the following forms. In an inserted Rsabha-carita ; kucchimsi,
159, 16 ; ulloyamsi, 161, 15 ; paramsi, 167, 17. An AMg-phrase : kucchimsi puttattde
. . . 76, 13 ; 91, 2h Besides 246, 27, manamsi (in a doubtful passage), 150, 20, tamsi
veldya (grammatically wrong !), 36, 25, sohanamsi, and 147, 16, etamsi (v. 1., etammi).
A SPECIMEN OF AECHAIC JAINA-MAHAEASTRI
329
sing, of feminine A-stems in -dya, but the locative in -dya and -dyam
{= Skt. -dydm !) ; similarly we have of feminine I-stems an inst. gen.
sing, in -iyd, but the loc. may end in -iyd or -iyam. Of these locatives
in -dyam and -iyam, the Pkt counterparts are preserved in the
following passages of : 280, 20, pavitthe ya dinayare, virattdyam
samjhdyam ; 280, 27, uttardyam disdyam ; 323, 18, uttardyam sedhiyam ;
310, 22, uttardyam sedhiya ; 312, 8, dattayam (sc. kanndyam) na
pabhavai sayano rdyd vd ; 210, 1, rdtyam (“ at night ”) ; 310, 24,
Pahhankardyarn nayariyam.
Feminine locatives in -dyam and -iyam have hitherto not been
known to occur in any literary Pkt dialect. Their genuineness is,
however, corroborated not only by the corresponding Pah forms, but
also by one sohtary form which Pischel (§ 388) quotes from a PaUava
Grant : “ Loc. sing. Apittiyarn (6, 37), i.e. Apittiyam = Apittydm,
i.e. the Pali form.” Now if -dyam and -tyam are genuine archaisms, it is
obvious that -dya and dya must also be recognized as such — which
in the case of the dative in -dya has, of course, never been doubted.
We have thus throughout an archaic form in -ya, a normal one in -e,
and we may therefore ask ourselves whether we have not here to do
with a uniform and purely phonetic development of final -ya after a
long vowel into -e. This would remove the difficulties with which the
explanation of the dat. sing, in -de is beset (cf. P, § 364), and would
make it unnecessary to have recomse to the Brahmana form mdldyai
for the explanation of Pkt mdlde (P, § 375). But whether this
be accepted or not, it is at least obvious that the substitution of a dative
in -de for that in -dya cannot be separated from the substitution of fem.
-de, -le for -dya, -lya ; if both changes are not the result of the same
phonetic law the former must be due to analogy with the latter.
One of the most interesting and remarkable archaic forms preserved
in Vh is an abl. sing. m. n. in -ain. I have noted it only thrice : 6, 13,
tao nissaranto diikkha-tmramm = tato nirasarisyad duhkha-
marandt, '* then he would have escaped a miserable death ” ; 146, 29,
1 The use of the present participle as conditional, taught by Hcmacandra (lii,
180; cf. Jacobi, BAan'-sa/fn A'aAa, § 35. In JErz no instance occurs, cf. JErz, § 112).
is remarkably frequent in Vh. I hare noted no less than nineteen cases, of which a few
may be quoted here : 149, 11, jai janaiilo, na ento ! " Had I knoBui (this), I should
not have come (with you) ! ” 126, 6. Juf em raggko honto, to padiynm mamam langhento —
na esa vaggho, “ If this were a tiger he would attack me who has fallen down — this is
no tio-er ! ” 1 10, 25, . . . kim puna tiiniain si annesim sahintao, “ (I should loye to tell
you), but you would tell others ' ” 120, 14, aham jai pamdena viggao honto. to mi
bandham jmrento, “ Had I stepped out by an oversight I should have been arrested.”
22S, 25. jai n na intitise ifi rhiam na damsenti, to mi i-ivanno honto, Had you not come
VOL. vni. p.cRTS 2 .vNii 3. 22
330
L. ALSDORF —
pawaya-lcandaram viniggayd — parvaia-kandardd vinirgatau\ 227, 24,
cukkd si dydram = bhrastdsy dcdrdt. There can be no doubt that the
forms in -am are ablatives : none of the verbs nihsarati, vinirgacchati,
cukkai — bhrasyati could be constructed with any other case. Now
an abl. sing. m. n. in -am may at first sight appear rather strange and
doubtful. Yet the development of final -at into -am is in perfect
accordance with the phonetic laws of Pkt, cf. P, §§ 339 ; 75 ; 114 ; 181.
Pischel says in § 114 : “ AMg sakkham = sdksdt (He i, 24 ; Uttar.
116, 370 ; Ovav.), besides S. sakkhd (Mallikam., 190. 19). AMg hettham
besides AMg JM hetthd (§ 107) is an accusative besides an ablative ;
the same may be the case with sakkham.” The three forms preserved
in Vh enable us to decide with absolute certaintv that sakkham and
heUhani are regular ablatives going directly back to sdksdt and
adhahsthdt. Their preservation is doubtless due to the fact that they
were used as adverbs and therefore no longer felt to be ablatives.
An additional proof is once again furnished by the Pali, where the
abl. sing, in -am also exists, though it is unknown to the official
grammar.! Of e.g. the well-known stanza Dhammapada 314 ( = Sarny,
^ik., vol. i. p. 49) : akatani diikkatani seyyo, pacchd tapati dukkatam j
katani ca sukatam seyyo, yam katvd ndnutappati, the first pada of which
is quite correctly rendered in the Vddnavarga (xxix, v. 53. “ B 41 ”
in Pischel s edition) by : akrtayi kukrtac chreyah ... It is clear that in
the 3rd pada, too. we have to translate krtac ca sukrtam sreyah. . . .
Or cf. Jataka 458, 13d ; knttha-Uhito para-lokarn na bhdye ? “ Being
in which condition need he not fear the other world ? " But it is well
known that the verb bhl is never constructed with the accusative,
but regularly with the ablative, exceptionally with the genitive.
or had you not shown me her figure, I should have died.” 36, 11, jai h piyn jivanlo,
tumam va uattha kusalo tionto. to na pin erisa-xirip. hhai/nnam hunto pram aimghadaga-
tiya-rauHa-camira-racckamuhpsu umhhnUo vih.arejja ' If your father were alive,
or if you were skilled in the .science of archery, this fellow" uoulil not cnjov siicli
prosperity nor would he thus roam about sportivelv in . . .” Further instane'es will
be found in Vh 13, 26 ; 16,2.5; 51,21; 71,22; 1,3.7,16; 137.2,21; 106 18 (read-
devam si . .) ; 168, 14 ; 169, 2 ; 228, 1 ; 248, 19.-The last of the .sentences just quoted
has already shown that for the expression of the irrpalia the ojitativo mav he
used as well. Other instances of this occur, c.g. 17, 4 ; 109 o . j-j-, -jq . jg
The two conditional periods 16, 2.5 f., and 17. 4. are both clear esprmplL’of'the'modua
irrenbi. \et in the first case the present part, is used, in the second the optative It is
very interesting to notice that Hemaeandra in his rendering of the storv in question
(ParisiPtaparian. i, 46 ff.) translates the present participles by conditionals (i 601
but the optatives by optatives (i, 72). ’ ^
> I owe the knowledge of the Pali ablative in -am and of the passages quoted below
to Professor Luders of Berlin. ' ^
A SPECIMEN OF ARCHAIC JAINA-MAHARASTRI
331
The verse Dhammapada 201 : jayam veram pasavati is usually
interpreted : “ He who conquers (jayan !) creates (prasavati) hatred.”
But a much more natural interpretation, corresponding much better
to the last pada “ Mtvd jaya-parajayam ”, is : jaydd vairam prasravati,
“ from victory arises hatred.” There can also be no doubt that
Dhammapada 49 : yathdpi bhamaro puppham vatum-gandham
ahethayamjpaleti rasam dddya, evamgdme muni care must be translated :
“ And as the bee, having taken the juice, flies awav from the flower
{puspdt !) without damaging its colour and smell, even so . . ." It
can be shown that in the dialect from which the Pali scriptures were
translated the abl. in -am was more frequent. But as in AMg and JM,
only those forms are preserved which had the good fortune of being
either overlooked or misunderstood.
A few pronominal forms also deserve to be mentioned. In Vh
mayd = Skt. mayd is the most usual form of the inst. sing, of aham ;
it occurs several hundreds of times. This form is nowhere even
mentioned in P, though it had been duly recorded by Jacobi (JErz,
§ 43). It seems, however, that in JErz it occurs only once (10, 1),
so that Pischel may have regarded this solitary form as a Sanskritism.
In the acc. sing. Pischel restricts the use of me to AMg. of te to AMg,
Mg, but we find the acc. me Vh 43, 26 ; 44, 8 ; 105, 18 ; 140, 7 ; etc.,
acc. te 65, 11 ; 81, 29 ; 84, 7 ; 87, 30 ; 221. 20. In the loc. sing.
Pischel knows »!af in ^ only : it occurs Vh 150.25; 241.13. Besides,
we read Vh 282, 21, the form mam-amkt. which is wanting in P,
According to P, §§ 419 ; 422, the plural forms ne and bhe are used as
follows : ne, acc. 51, AMg, inst. gen. only A5Ig ; bhe. nom. taught
by one grammarian (Canda), acc. inst. only A5Ig ; gen. A5Ig and JM.
In 5^h we find : ne, acc. 121. 13 ; 134, 27 ; inst. 233, 27 ; gen. 43, 3 ;
70, 24 ; 73, 23 ; 94, 16 et passim, besides acc. pi. nam (read ne ?),
70, 23 ; 230, 23, bhe, nom. 99. 26 ; 125, 13 ; acc. 118, 4, 8 : 153, 23 ;
367, 17; inst. 101, 4; 108, 3; 112, 10; 115, 26: etc.: gen. in
common use. According to P. § 422, the gen. pi. co = vah is used in
51, S, and the Pallava Grants, but not traceable in other dialects :
it occurs Vh 211, 27 ; 224. 3 ; .351, 2. Besides, an unmistakable
nominative vo is found 5"h 88. 21.
bhe is explained by Pischel as the result of a weakening of tubbhe,
due to ab.sence of accent (§ 422). This explanation is confirmed by
a hitherto unknown form found in Vh ; a genitive mhe bears exactly
the same proportion to tumhe as bhe to tubbhe. I have therefore no
doubt as to its genuinenesss, even though it occurs only four times,
332
L. ALSDORF —
within thirty consecutive lines of the text ; 213, 8, ahavd to [tubbhe^ ^
bhuena Icenai mhe ^ urnhd ® bojjd ? Or should you have got fever
through some demon ? " 213, 15 : na mhe urnhd sanrassa, “ You have
no fever ! ” 213, 18 : kerisam mhe * sanrassa ? “ How is your health
(lit. the state of your body) ? ” 214, 2 : ma)m pasderia mhe ® kajja-
siddhi. The readings of the MSS. plainly show that mhe was unfamihar
or even unintelligible to the scribes so that they tried to elimiaate
it — which for us is an additional proof of its correctness and
authenticity.
It would be easy to add a long list of other forms, words, etc.,
occurring in \'Ti which are either completely wanting in P or attributed
by Pischel to AMg only. A brief enumeration of at least some of them
may be given in conclusion.
dyd = dtman (§ 88, also JS ddd), 130, 16 ; -tra after long vowel
becomes -ya (§ 87) ; gdya = gdtra, 73, 12 ; 196, 10 ; 320, 14 ; 328, 21 ;
gcyya = gotra, 159, 14 ; I'idiyd = vidyd (not in P), 88, 14 ; nitiyarn
= nityani (not in P), 178, 27 ; ahe = adhas (§ 345), 152, 3 ; 155, 18 ;
ahe-loga, 159, 20 ; uydhu (§ 85), 59, 9 ; 151, 13 ; uddhu, 12, 6 ; 37, 20 ;
47, 27 ; 126, 23 ; 135, 28 ; 142, 30 ; voc. rdyam ! (§ 399, not in JErz),
128, 30 ; 131, 9 ; 234, 4, 9 ; 244, 13, et passim ; nom. pi. bahave
(§ 380, “ in JM probably wrong for bahavo ") 52, 19 ; 234, 27 ; 310,
22 ; 323, 26 ; nom. pi. gdvo = gdvah (not in P, § 393, but cf. Geiger,
Pali, § 88), 181, 28 ; gdo = gdvah (§ 393), 182, 3 ; kdladhammund
samjutta (§ 404, p. 284 bottom), 75, 25 ; 284, 20, 25 ; 286, 20 ; 287, 4 ;
298, 15 ; 299, 9 ; 304, 4, 5, 7, 9. Absolutives ; in -ttdnam (§ 583) :
bkarettdnam, 53, 27 ; pakkhdlettdnam, 247, 20 ; uddissa = uddisya
(not P, § 590), 14, 8 ; paducca (§ 590), 5, 17 ; 10. 23 ; 280, 31 ; 311, 13 ;
342, 16 ; 360, 26 ; pappa = prdpya (§ 591 AMg and JS), 235, 5 ;
dydya = dddya (§ 591 AMg and JS) 163, 20.
Ml has presented us a picture of JM materially different from that
which is familiar to us from P and JErz. Its main features are : first,
an even much closer affinity to AMg than that prevailing between
ordinary JM and AMg (cf. P § 20) — we are markedly nearer the time
when JJI came first into being as an individual dialect, distinct from
AMg. Secondly a great number of archaisms and other interesting
^ Inserted, because the following mhp. had lieeome unintelligible ’
2 Only one MS. reads thus, all others have ftp..
^ The editors print u<j<jd which gives no t.en'.e ; cf. the next passage.
* One reads thus, the others have he.
* All M>SS. but one read mhi.
A SPECIMEN OP ARCHAIC JAINA-MAHAEASTRI
333
peculiarities whicii not only in themselves are valuable additions to
our knowledge, but also help to elucidate many a dark point of Pkt
grammar and linguistic history. Ultimately the great problems of
the origin and true character of AMg and JM, of their development,
and their relations to each other as well as to the other Pkt dialects, the
Pah, etc., are raised anew and demand new answers.^ A discussion
of them is, however, quite beyond the compass of the present paper,
the aim of which has merely been to make known the new material
supplied by and to direct the attention of fellow Indologists to a
field of research which in my opinion badly needs new tilling.
^ The conclusions arrived at by Jacobi in his paper “ Uber das Prakrit in der
Erzahlungsliteratur der Jainas ” {RiviMa degli Studi Orientnli, m08-P, pp, 231 ff.)
will have to be modified.
Yazdi
By H. W. Bailey
TYTHILE I was in Isfahan at the end of June, 1932, a young
* * ZarduAti of Yazd, Jam^d Khoda-morad Ra^Idl, came to
see me before going on to his school in Shiraz. We met four times,
and before leaving he requested an acquaintance of his, a young
man, Jamshid Qobad, who was visiting Isfahiin, to come to see me.
Qobad's home was with his parents in Teheran, so that he had little
opportunity to use this dialect. In August of the same year I was
able to \dsit Yazd for one week, and there found a young Zardushti
school-teacher, ArdesMr Mehraban Mazkuri, born in the neighbouring
village of Maryabad (or Moriabad). We met several times. He gave
me the following list of eighteen villages where Zardu^tls lived :
yoramsah, yeirabad, ahrestan, zeinabad, moriabad, hasanabad,
nosratabad, nasriabad, elabad, hosseinabad, mehdiabad, aliabad,
mazree ardesir, kQce bfuk, yasemabad, cem, mobareke, taft. In
yasemabad he said that only Zardu^tls were found. I was able to
visit Taft and to spend a short time there, speaking with some
Zardushtls. The words I was able to write down showed little difference
from the language used in Yazd, Mazkuri, however, gave me many
differences between town and village pronunciation, and a few
differences of vocabulary. It would be useful to have a full linguistic
study of the whole district.
The sixteen tales which follow all suffer to some extent by being
translations or adaptations either from Persian or from English. The
translators have not always freed themselves from the Persian of
the original.
In the pronunciation, 5 and u were at times hard to distinguish.
As to se (indicated by a in the texts) it seems to have been the usual
sound for Ra^Idl and Qobad, whereas a ‘ was commoner in Mazkuri's
pronunciation. Length of vowel varied also.
Morphology
Pronoun
Nom. independent : me, ta, u. In, Ine, mo, semo, smo, le, Isun.
Possessiv^e suffixed : om ot os mo do so.
Obj. infixed ; -m-, me-, om, me I mo mu
ta ta de j smo semu
os
le
^ a is used for the sound [a].
336
H. W. BAILEY —
Obj. with 6, vu “wish, must”.
me de se mo do so.
Agent with Pret. :
em mem
od tad
mo momo
do semodo
os UOS, US
SO leso, Jstinso
Agent with Perf. and Pluperf. ;
ml dl si
mo do so
mi di si
mo do so
Verb
Present and Preterite bases are distinguished. Intransitive
Preterite is expressed by the
-t participle and
verb substantive ;
Transitive Preterite by -t participle and agent form of pronoun.
Verb Substantive.
Pres. Indicative
Copida.
Pres. Conjunctive.
me he
e
be, vebe
ta hi
i
u ha
on, 6, un, un
mo him
him
semo hid
hid
isun hend
en
Verbal Paradigm.
Pres. Indicative.
Negative.
Pres. Conjunctive.
mekre -kre -ekre
me nakrime
-vekre
takri -ekri
ta nakriye
-vekri
• , , f rnimd
ukra -ekra 1 ,,
fsejit
u nakre
-vekra
radkrim -ekrim
mo nakrime
-vekrim
imokrid -ekrid
semo nakrite
-vekrid
iekren -ekren
isun nakreme
-vekren
isunekren
Imperfect : -ka -ka
meka mOka
According to Mazkuri ka was the
deka dOka
pronunciation of the town, ka of
sekn soka
the village.
Transitive Preterite.
Negative.
memka omka
memnaka
-omnaka
tadka odka
tadnaka
YAZDI
337
uoska oska
usnaka
momoka moka
momonaka
semodoka doka
semodonaka
isunsOka sCka
Isun sonaka
Transitive Perfect.
Negative
me mikarta
me minekarta etc.
ta dikarta
u sikarta
momokarta
smodokarta
iesokarta
Transitive Pluperfect.
Negative.
me mikartaba
me minekartaba
di-, si-, mo-, do-, so-
Transitive Pluperfect II (karde bude am)
me mikartebda
Negative.
me minekartebda
di-, si-, mo, -do-, so-
Intransitive Preterite.
me ras6d-e
boi, boie
ta ras6d-I
boi
u ras6d
ba, bo, but
mo ras5d-Im
boim
semo rasod-Id
bold
Isun rasod -an,
-and boiend
Intransitive Pluperfect.
bedaboie
bedabolm
bedaboT
bedabold
bedaba
bedaboiend
etc.
etc.
Imperative. Negative.
2 Sg. maku
2 PI. makud
3 PI. nakren
na- in place of ma- in the 2 pers. was censured by Mazkurl.
Infinitive.
(1) -tvun, -tun, -dun, -vun.
(2) -t infinitive.
va;^te dovrfrt
va;^te davOrno
338
H. W. BAILEY—
Frequent with veo “ must ” ; os-veu ced “ must gather ” ;
os-veo asnoft must hear ’’ ; veo sa “ must go ” ; om-vau pars5d
“ I must ask ” ; nautoma, naotoma must not come ” ; naosa
“ must not go ” ; naoka " must not make
Passive.
Expressed by the verbs b6d\’un and omdvun.
venodabo : mo guve venodabo “ this ball was thrown
mart oma " was broken ” ; mart! 6mdab6 “ was broken ” ;
menesti 6mda “ was iv-ritten ” ; meneste tod is written ” ; nodi
Omdaba was placed
List of Verbs
amorz-
blamorzi, memblamorzun.
“pardon”
arotvdn
meare taari, Ine ard, arlm, -s ort.
“ grind ”
arvastvun, arvastvun, arv'dstun
“ leap ”
mene areveza mijehad ”.
arvast, drvhst.
asnoftvun. asn6pt\'iin .....
“ hear ”
me snave, me suave.
memasnoft, memasnoft (tad-, ines-),
osasnoft.
soasnoft.
Inf. asnoft : os veo asnoft.
bartvQn
mebre mebre bren. Conj. me5 gi vebre.
membdrt, sobd, iesubdrt.
“ bear ”
Inf. osvau barte.
bastvdn
me benne, meubenne. Imperat. maben.
membdst, sobast.
“ bind ”
ba;^s6d
me vebayse.
memba;;^sod, osba;^s6d.
“ bestow ”
bodvun,
bodvdn ......
Pres. 3 Sg. but. Conj. 3 Sg. vebe, bfld.
“ become ”
Pret. boi, boie. boi, ba. bo, but, boim, bold, boiend.
Pluperf. bedaboie, bedaboe, bedaboi,
bedaboim, bedabold, bedaboiend.
Imperat. be, vebe.
bedaba,
YAZDI
339
brldvfln, burid\iin .
. (1) “cut
”, (2) “kill”
me vebrine. 3 Sg. brina.
membrid.
car- .....
“ graze ”
Pres. 3 Plur. cdren.
casnodvQn ....
“ join ”
me casne, memcasnOd.
cedvun .....
(1) “gather’’,
(2) “weave”
me vecine.
memced, osced.
Inf. osveu ced.
cerzod . . . .
“ complain ”
3 Sg. pret.
cos- .....
“ taste ”
Imperat. vecos.
daridvQn ....
. “ tear ”
me vMrine, memdrfd.
dasvun .....
. “sew”
me vedarze, memda.st.
davodvtin ....
. “run”
3 Plur. daven.
3 Plur. davdden, me . . .
davortviin ....
davode.
“ turn ”
me vedre.
me dvdrte. 3 Sg. dav^rt.
Inf. va;\^te vcdrit.
va;^te devart.
davorna .....
“ turn ” (tr.)
Inf. va;^te davOrna.
did .....
. “ see ”
Pres, see vin-.
memdid momodid omdid modid
taddid semododid oddid dodid
lesOdid osdid sodid
me mididaba. dldldaba, sididaba, momodidaba,
semododidaba, isunsodidaba.
Inf. oni veo dide.
dodvun ......... “ give ”
mgte, me . . . ete, me . . . ate, pete “ I give
3 Sg. eta.
340
H. W. BAILEY —
Neg. 2 Sg. natle.
Conj. 3 PI. aden, aten.
Imperat. 2 Sg. adem (be-nie “ to me ”).
Pret. memdod, omdod, osdod, se sar do, tot, ostod.
Perf. -s dode, iesomeidoda “did not give me ”.
dortvun ......... “ have ”
me ddre. Neg. nadore. 2 Sg. dorl, dori, dOri. 3 Sg.
d6ra, dora. 1 PI. dOrim, dorfm. 3 PI. doren.
Pret. memdost. osdost, -stos, -sdost, -sdort.
dosod’^un ........
“ milk ”
me vedusen, memdosod.
gasvun ........
. “ bite ”
me geze, mesgast.
gertvun ........
“ turn ”
me gerte, me gertode.
grafvun, grdftdn .......
“ take ”
Pres, me pnore. me zem gore “ I learn ”.
Put. gdre.
Pret. omgrnft, meme zem grapt, zem graft,
henddvun ........
. “ put ”
Village pron. = town pron. nodvfln.
heridvun ........
. “buy”
me herine, me v4rine.
memhrid, seherlt.
Imperat. verin.
hermondvQn .......
“ command ”
me vehermone, memhermonid.
hrotvQn .........
. “ sell ”
me vgrose, memh'rSt.
jom- .........
“ move ”
3 Sg. jome.
jomnod ........
“ move "
me Jomne, me Jomnode.
kapt .........
. “ fall ”
Pres, tebe ka “ falls ”, tebe ke “ I fall
Pret. krfft, kapt.
Pluperf. kapteba.
kartvun. See Paradigm above ....
“ make ”
Inf. ka (naoka, omveuka).
Imperative 2 Sg. maku, nakud, nakude.
YAZDI
341
kast ......... “ plant ”
koren, sokast (village pron.).
sokast (town pron.).
kenodvun dig ”
me vekene memkendd.
kesodvun ........ “ draw ”
me kese.
memkesod oskesdd kesOd.
kostvum, kosvun. . . (1) “ throw ”, (2) “ strike ”, “ shoot ”
me bekude.
me me gulkos “ I deceive ”.
memkost -skost me me gulkost.
Perf. 3 Sg. kdste.
kust- “kill”
kusim.
kuste bdien.
lest ........ . “ lick ”
Pres. 3 Sg. lesa, velesa.
Prct. -s leSt.
mart ......... “break”
1 Sg. emare, me vemare. 3 Sg. mdrd,
memmart, ommdrt, osmdrt.
mdrtl dmdabo, mdrtoma “ was broken ”.
Participle mdria “ broken ”.
marf'dn die ”
me vemre.
Pret. me marte, mdrte, ta mart!, u mart, mo martim,
semo martid, Isun inartan, marten.
Partic. mdrta “ dead ”.
mekdd ......... “ suck ”
ine meka, osmekOd.
menodvun . . . . . . . . “ remain ”
me mine. 3 Sg. mina, mlnid. 2 PI. mlmd. 3 PI.
menan.
me menod.
v6 menud “was exhausted”.
mnesodvun ........ “ write "
me vemncse.
me mnest, osmenes. Perf. 2 Sg. dl menesta.
Pass. mene.sti onida, meneste tod.
342
H. W. BAILEY —
molldvto mb ”
me mole, memolid.
nasvQn nasvun . . . . . . . . “ sit ”
me mge, me diiige, ine nfga, me nSste. 3 Sg. ndst, nast.
Pluperf. naita boien.
Imperat. 2 Sg. ma-nfg. 2 PI. flnlgid.
neftvun ........ “ send ”
me venive.
me . . . n6f, de . . . ngf, se . . . nift, osngpte, memneft.
me mi neptaba (-neft-) slneptaba.
uslnepta.
nodvdn ; cf. benodvun .......“ put ”
me tu motor nenom “ put me into the motor ”.
nod, sonOd, sonfid, O.snode.
nodi, omdaba.
omvun, omdvun ....... “ come ’’
toe toi tod, tud tOfm toid tOiend
Pret. me onioye mo omOyim Neg. nei6mda.
ta om6yi .semo omOyid
u oma isun Omoyend.
Pluperf. Omdaba.
Participle Omdd.
Imperat. biOr, bIG, biOid.
Inf. naOtOma.
ortvfln “ bring ”
me tore, tOre, me tOre, etOren.
Conj. 3 Sg. biOra.
Pret. osOrt.
Inf. osvau tOrte.
parOdvQn . . . . . . ,
me pare. 3 PI. parOden.
parsOdvun . . . . . . . “ ”
me parse, memeparsOd, ospars6d.
'-cook”
me vepese, mempa;^!-
P«niOd “ understand "
pahmend. vcfahme. Conj. 3 PI. vefahmen.
Pret. pdmOd.
Picod ... “twist"
me vepice, me plcode.
YAZDI
343
pos- ........
ove pOsd “ he sprinkles water ”.
“ sprinkle ”
prontvdn .......
me aprone, me pnore, me pnOre, me prone.
3 -Sg. -e pnOra.
mempront, ospnort, pnort pront,
ospron, osopront.
Perf. Neg. 3 Sg. naprSnte.
. “take”
ranjodvun .......
me ranje, me ranjode.
“ be troubled ”
ranjnod .......
me ranjne, memranjnod.
“ trouble ”
rasodvfln .......
me rase, rasim. Conj. 3 Sg. naresi.
Pret. see verbal paradigm above.
“ reach ”
retvQn ........
me rije, me Qrlje. 3 Sg. rfja.
memret.
“ pour ”
roft ..... . ,
ine ruvd.
osroft.
“ sweep ”
ronodvun .......
me rone, memronod.
“drive”
sat ........
senjodvun, me senje.
-s sat.
“ weigh ”
se]- ........
3 Sg. sejid, s6ji, sejit, seja, sezid.
“ burn ” (intr.)
sotvun ........
me vesoje. 2 Sg. sojl. Conj. 3 Sg. vcsuja.
memsOt.
“ make ”
su|n6d ........
“ burn ” (tr.)
me suzne, me vesuzne, Ine siizna. Conj. me vesuzne,
memsuznod, sujnod, sQznod, suzn6d.
suin- ........
me suine. 3 Sg. sQina.
. “ burn ” (tr.)
salodun .......
“ to limp ”
semordvun .......
“ count ”
me vesmore, memsniord.
344
H. T,y. BAILEY —
senosodvun ......
“ know
me beindse, memsenasod.
sadedvun ......
. “take
mesne, tSsm', sina, inesna, esnd, asnen, mdsnlm,
smOsnid, Isunesnen.
Pret. om saded, odsaded, ossaded, -s sided, m6m6-
saded, smodosded, iesosded, sosted.
Perf. me misdeda, ta disdeda, sisdeda, momosdedfi, smodos-
deda, Isdnsodeda.
Inf. osvau sdede.
snost .......
-6s snost, -6s sandzd.
. “put”
vastvun .....
me veze.
me vaste.
“ run ”, “ run awav ”
venodvdn ......
“throw”
1 Sg. evene, ine vena. Conj. 3 PI.
venen.
Pret. 3 Sg. ven, osven, ven6d.
3 PI. sovendd.
Pass, venodabb.
1 PI. venodim.
me vlje, uvije.
. “ sift ”
vin- ••.....
. “ see ”
vevine, vevini, vevTna, vevimm, semo, vevlaid, Isun
vevinend.
Imperat. vevin.
■viosvun
Pres. 6, vu, ii, vud, ode, veo, veu.
Neg. navate.
Pret. vevi6ste, sevlost, devio.st.
. “ wish ”, “ must ”
vistvun .......
me viste, me vTstode.
bari me viste “ stood by me ”,
“ stand ”
vizdrt (village = town davdrt)
Inf. va;^te vezrit.
“ turn ”
vazorna ...
Inf. va^te vazdrna.
. “ turn ” (tr.)
v6ptvun
<< „ >5
me vope, memvopt.
weave
YAZDI
345
Yortvun ......... ‘’ bring”
me biore, bldre.
memvOrt, osvurdt.
me mi vorta.
Imperat. bior.
votvTin . . . . . . . . . . “ say ”
3 Sg. vaja, vaja, vaje, vevaja. 1 Sg. me vevaje.
2 Sg. vaji. 1 PI. navajime.
Conj. bevaja.
Pret. memvot, -dv6d, usvod, sevod.
Perf. sevota, sOvota.
Imperat. beva, veva. nava, mava.
Xart^un “ eat ”
me va;)^re. Conj. 1 Sg. ve_;^re. Neg. na;^re. 3 Sg.
ve;('ra, be;)^ra. 3 PI. va;^ren.
Fut. mexre.
Pret. mem;;^art, os^a.
Imperat. td ve^o.
xend “ call ”
me xine ni6 xiniai
ta xiiii smo x^-ald
u xina ie
Conj. me vu vexine.
Pret. moxend.
Xofvun, xbfthn .....
“ sleep ’
me xopte. 1 Sg. xbptude.
Phiperf. 3 Sg. xbpludabo.
zodvun .......
" be born '
Pres. 3 pi. zoyen.
zSnodvun ......
“ bear children ’
me zOne. memzondd.
zonodvun ......
" know '
me zone. 2 Sg. zOnl. 3 Sg. ezona, memzonod.
Perf. Neg. neizonoda.
Vocabulary
ado, with dr (village), with
adovoi, with ’ar’ar, braying
adovoi, ddovdi, adSvos arbob, arbub, master
dldi, money aro, from
VOL. Vin. PARTS 2 AND 3.
23
346
H. W. BAILEY
asb, horse
asbe, asp-davuni
asta, bone
avaz, return, substitute
babu, father
bad, bad
bad, after
banafs, violet
brfr, bar, outside
be, other
be, quince
bedar, bedere, bder, father
benOha, began
benoska, benuska, began
beridn, a roast
bidi, again
bldor, awake
blguna, stranger
bland, high
boy, garden
boyviin, gardener
bold, above
b6na, pretext
bor, load
bri, for
bumas, grandfather
bur, time
hesbur, hesbure, never
ya buri be ]I, again
cakakara
v6 cakakara, water is dropping
cap, left
carbi (village pron. carbi) fat
capta, polo-stick
ce, what
cem, eye
cen, cend, centa, cem (bur). Pcrs,
cand
cerul, a pasture
camus, shoe
CO. Pers. cun
coyoli, complaint
core, means
coy, sound to drive off dogs
cfiyl o loki, in good condition
cum, thing
CUV, wood
dast, dast, das, hand
dasti gora, grapes
davaj, bed
ddvrl, around. Pers. dor-e
de-, into
de, vnllage
dend, wasp
deres, derest, correct
devar, plur. devarQn, brother ;
daveri mas, elder brother ; be-
dverl y(7 to one's own brother ;
devarona, adj., brother's
devQte, inkstand
davurodavur. Pers. barabar
dl, Pers. ham, also
dikartvun, to find
sodika, they found
dir. 1 Pers. dur, far. 2 Pers. dir,
late
doy. Pers. day
dom, tail
dombol, dumbol, behind
dOnI, world
doz, dozz-, thief
dravs, banner
drayt, derayt, tree
drujna, window
drdv, false
emoma, turban ; Pers. ’amamah
emru|, emrd, to-day
YAZDI
eskam, belly
ezm, permission
galihem kartvun, to mix
gall mol, neck
gap, words
gap kartvun “ to speak ” ;
gape-ro, gafe-ro Acc.
gdrm, hot
gdv, gau, (1) down, (2) deep
gas. Pers. negahdarl kardan
genniim, wheat
genz, room
gerd, round
gir, girl, capture
gSmbur. Pers. gahanbar, festival
gor. Pers. gabr ; used of the
Zardusti
gorba, cat
gori, rujgoriaka, ruzgoriyaka,
greeting during the day.
gos, ear
gos se gaf k(T, he heeded
gose, corner
gu, gQve, ball
gu, ox. Pers. gav
gure xcr- gura ;;^er, wild ass
gflra, gora, unripe grapes
gurg, wolf
hama, hama, all
hardo-ro, harduta, both
haruja, every day
heci, none
hendl, water-melon
herdfl, to-morrow
hes, his, plough
hes, none
hes va;^t, never
heske, no one
347
best, mill(?)
heze, hezze, yesterday
hezvun, tongue
bo zvun, bo zfm, with tongue
hoderi asb, horse-keeper
hoi. Pers. hal
hosa, ear of corn
hosk, xoSk, dry
Ja, barley
|edf, gum
Jegar (village pron. Je;\;ar). liver
JelaverT. Pers. ru-be-ru
|evun, fine, beautiful
jola, yoke
kad, yadd. Pers. qadd
kaleSca, jar
kappa karf^un, to tear to pieces
kapal. Pers. sarin
kasogi. childhood
kasok, kasuk, kasoye, small
ke, who
kemogi, a little
kezi (village, keza). Pers. kadu,
gourd
ki, that
kie. Pers. kucah, street
klop, cheek
kluft, thick
k6i, kd}-!!, where ; de-koi, to
what place
kOnd, old
kor, kur, work
kur, blind
korbun. Pers. qurban
loyar, thin
loki, see cuyi
lonj, lip
lop, mouth ( ? J
348
H. W. BAILEY
maid, kind of grape
mas, mas, great
massere, greater
masln, motor car
md;:^e, lost
mehr, friendliness
meimun, (1) guest, (2) monkey
memas, grandmother
memu, mother
mer, mother
merdog, man
mes, sheep
meyune, middle
mid, hair
mir, husband
mire ya, one's oi\ti husband
mo, this
mol, neck
momne, now
mone, mdne. here
mor, seal. Pers. muhr
m6sg(e), mouse
mosolvun, Musuhuan
nafahml, lack of understanding
nam, wet
ndrm, soft
narmi,
nast, uneducated
naxa.s, unhappy
nayod, lentil
ne, now
nebdi, now. Pers. hala digar
nemak, salt
nimrfl, mid-day
num, name
numa. Pers. namah
nur, pomegranate
6. Pers. an
6, v6, water
6, v6, and
obed, rehgious man
obf, blue
ohessa, softly
ohosa, lamenting
ohonOla, lamenting
on. Pers. an
one, there
ort, flour
osurl, a headstall
ostor, hostor, camel
otes, fire
otrafe, otrape, surroundings
6;)^ari kor, finally
pall, side
pdn, wide
parande, bird
parr, wing
passln, evening
peiom, message
pesOr, pressure
pesseherdQ, day after to-morrow
pi, fat
pir, old
pisi, to
p6, pu, foot
por, por, par, fub
porzur, strong
por, son
porlke, son
porog, pi. porogun, son, bov
poska, put on one's back ?
prli, down. Pers. furud. pru so,
furud raft
ra, in, id, road
rauza. Pers. rauzah
raz, grapes
res, beard
YAZDI
349
ri, upon. Pers. ru
se ri, upon
ro, road
roy, vein
rost, true
rosvo, cheat
rubas, fox
ru], day
ya ruji be, another day
ruj goriaka, ruzgoriydka, greeting
during the day
ruja, a fast
ruva, soft skin (Pers. sifale narm)
(of wheat)
sabo, morning
sab6;^eir, saboyeira, saboyara
greeting in the morning
sag, dog
sara bolfli. Pers. gardanah
sarbozi pir, an old soldier
sarosiv, head down
sarvon, leader of caravan
sauzl, greenness
savz, green
sejen, sejen, needle
senjed. Pers. sinjad
serv, cypress-tree
sTo, black
sirat. Pers. surat
soat, hour
soni, hoof
sor, red
su — u-su, at that time
svid, sevid, white
sabeyara, sav6;^eir, sav6;(eira,
greeting at night
sdr, sahr, city (village pron. sar,
sar)
sau’', sav, night
sa^, firm, hard
se, to
sekar, sugar
siv, down
se sM bor, under the load
sod, happy
suna, shoulder
tdl, bitter
tang, narrow
tar, fresh
tdsua, thirsty
tasnegi, thirst
tavila, stables
teki r«, small part of the way
tel, body
tenhd, alone
tep, tep krfft, he fell ; tebe ke,
I faU
tid, mulberry
tij, sharp
tolbpl, recompense
tong, vessel
torik, dark
tros, tros, sour
vacca, child
vaccegl, childhood
varf, snow
vdsnd, hungry
vdsnegi, hunger
vdter, vater, vatar, better
vayt, time
verrd, lamb
ves, more
vester, more
vi, vikar. Pers. bl-kar.
vi-poida. Pers. bifaidah
vid, willow
350
H. W. BAILEY —
viovun, desert
vir, memorv
vis-solegi, twenty years of age
v6, 6, water
vodem, man
vodim, almond
voj, voice, shout, noise
vokre, I open
vovizun — me vovizunka, I
suspend
vozod (village pron. ozOd), free
;)^ados, himself.
Xa, one's own
Pdaloni, desolation, ruin
;^am gardesi, winding road
xar, ass
;;^argeri, folly
;^as, well
Xashdl, happy
^(evar, sister
xda, house
^ib (village pron.), x^b, good
Xin, blood
Xis6y5m, relatives. Pers. x^s u
qaum
Xodo, xado, with
Xodo, Grod
xomus, silent
xordekra, he grinds
XO.sk, hosk, dry
Xrov, ruined. Pers x*‘'rab
Xub, good
Xunendegl, calling
zdrd, yellow
ze, bowstring
zem, memory. Pers. yad
zenda, alive
zendegl, life
zevin, earth
ziod, zidd, more
zur — me . . . zursekre, I press it
I
va ru]I ya x^rl yado yak hostor pise yak arbob kor eska. voske kor
eskardaba loyari bedaba. arbubos hardoro be-sahrd se sar do. mie
hem soend va yak cerdl sodika. va to cend vaxtl v6 u6 sauzl soya va hoi
omoyen. yak ruji yar v6] osdo.st o 'ar'ar oska. ho.stor bicora harci dod
oskost voje ya maku yadoinl pahniend va toen va mopronen va
dobOra bor mo bor ekren va dar mosibat movenen. vali yar yeili
yargeri-s ka va gose gap naka usvod ki yunendeglye bederom me
virl Oinda va me lii veylnc. dar on vayt be-hokme yodo yak yadomi
az otrape devart. sarvon vojos o.sa.^n5ft va domboll vOjos sa. yeili
ra nasa gl yak yam-gardcsl 6ma. osdid gl ya yar o o.storl cuyi 6 loki
dar mdne cdren. osdri harduta osprOn osvurdt va se sivl bor oskesod.
hostor 6tes be-del voji kasoye dosnum se yare dot. az nafahmiye yar
ddbora so girl venodlm. hostor dOyl zidd ke dar delos nodi omdaba
pisi yados sevSd dobara tolopi se vokreh doy o margl yar vaytos gi
bdd ya teki ra ki sa yar benoska salodun. sodid ki yar sail beda.
boros .sosted [va boros sosted] va sonOd sar bori ostorl bicora. hostor
YAZDI
351
I
Once an ass with a camel was working for a master. So much did
they work that they became thin. The master let them both loose in
the fields. They went and found a pasturage, and for some time they
drank the water and ate the herbage and came to a good condition.
One day the ass gave voice and brayed. Although the unfortunate
camel cried out, Stop braying, someone will hear, and come and take
us and again bring us to our burdens and throw us into misfortune, yet
the ass persisted in his folly and would not listen, but said, The bramng
of my father has come into my mind, and I wish to bray. At that time
by command of God someone passed In'. The caravan-leader heard
his voice, and followed his voice. He had not gone far when he came
to a winding way. He saw an ass and a camel grazing there in a fat
condition. He took hold of the headstalls of both, and brought them
and placed them under the loads. The camel, fire in heart, in a low
v'oice cursed the ass. Through the ass's stupidity we have again been
taken. The camel, with great indignation in his heart, said to himself,
352
H. W. BAILEY
ohessa pisi ;;^ados osvod bah bah ;^ub korl moka. bedi yak teki ra
ki sa yar vOmenud va dast o poyos sobast va soniid sar bori hostor
ceroki. b5r6s kesod. hostori bicora pIsi yados cerzOd va so to rasuden
be yak sara bolQi. injo saroslv vau\y6sti sa. hostor benOska arvdstun.
;;(ar benuska ohonOla kesodun ki mane tebe ke. hostor Jevubos osdod
ke de-vir etod ki narmi devod ;^unendeglye baderod de-virl Omda.
va me dl arvastuni merom me vlrl Omda. hostor arvast va yar se" gau
venOd. ;^ar ke az beri ostor gau kapt va asta-s mart oma va mart.
az mo hekayate mallm on nasihate yadomi ki gos nakrime vayti
mukopot rasim mesli ki yar rasOd. temQm ba.
See JRAS., 1897, 103 fif., and JRAS., 1932, 403-5. An ArdistanI
version is published in BSOS., vii (1935), 775-6.
II
yak podesOI ado nokaros ajami dar yaki kasti nastaboien. nokaros
liesvayt deryo si nedidaba o sayti kesti fekrosT nekardaba. ohosa
.sesuTui kardaba vo larz se badenos kapteba. harce mehrabunl soh6ka
orum naprdnte. podesoiro eisos monayyas va hes c6rei fekrosos naka
va yak tabibl dar mo kasti ba. podes6 ado tabibos v6d ta hes c6rei
zdn! osvdd ager hokm vekri me uro yomiis ekre. podiso osvdd yeili
me yase t6d. tabib osvod to nokar ro tu deryo venen. vayti kapt
tu deryo yota cend osya. pas mid 6sopr6nt so venod tu kesti. vaytiki
sovenod tu kesti sa v6 yak g6§e nast. podesO ro ajabos oma. in ce
hekmati ba ke tadka. osvOd cun nokarod mehnati yary bodvun se
neizonoda vo yadri salQmati kasti ro se fahmi neikartebda. bemcenun
yadri tenderesti yadOmi ezOna ki yak mosibati gir vekra.
Ill
podesoiro hama sOasnoft ke bekosvfine yak bicorei hokmos ka.
mo bicore dar holati nayas va bifekri bo zvflni por toyat gi-s tos
bepOdeso dosnQm ostod va badozes bepodeso sesuroka mesleki sovOta
harke dast az june xados besfira onciki dar del dOra bevaja. podeso
ado nokaros osparsod ke mo cici vaja. yaki az mi vazire ge hozer ba
osvod ei soheb mo vaje gi ta biy6d hokm dekOsteni uni karda. vaytiki
mo osasnOft az vei rahmos Oma vo az sare yinos devart. vaziri digar
ke zedde mo bicOre ba osvod porogOn jensi mOro na.snasene ke dar
hazrati podesoun joz berOsti sohbat kartvuni u podesoro dosnum
osvOd badozest osvdd pOdeso sar o rise tuhem kesod va osv’Od me az
YAZDI
353
I will requite him for this, when pain and death wiU befaU the ass !
When he had gone a short way the ass began to limp. They saw the
ass was limping. They took the load and put it upon the load of the
unfortimate camel. The camel said softly to himself, Ha, ha, we have
done a good work. Again, as he went a little way, the ass was worn
out. They bound him hand and foot and put him upon the camel's
load, so that he carried his load. The imfortunate camel lamented
to himself and went on till they came to a defile. Here it was necessary
to go with head down. The camel began to jump about. The ass
began to complain, I shall fall down. The camel replied, You will
remember tliat you said softly (?), The braying of your father had come
into your mind. And now I too remember the jumping about of my
mother. The camel jumped about and he threw the ass down. The
ass fell down from the camel and his hones were broken and he died.
From this tale it is clear if we do not pay heed to one, we shall
come to a time of recompense, as the ass came.
II
A king with a foreign servant was seated in a boat. The servant
had never seen the sea and greatly feared the boat. He began to weep
and his body trembled. Whatever kindness they did him, he was not
quieted. The king’s pleasure was spoiled and he could think of no
remedy. A physician was on the same boat. The king said to the
physician. Do you know any remedy ? He said. If you command, I
will make him silent. The king said, It will please me greatly. The
physician ordered to cast the servant into the sea. He fell into the
water and was immersed several times. Then they seized his hair and
brought him into the boat. AYhen they had brought him into the boat
he went and sat in a corner. The king was surprised. ^Vhat is this
wise thing you have done ? He said. Your servant had not known
the distress of being drowned and had not understood the safety of
the boat. As one knows the value of health when a calamity seizes
upon him.
Ill
Of a king it is related that he commanded to slay a certain
unfortunate. This unfortunate in his evil and desperate condition
with the powerful tongue that he had spoke ill of the king and began
to abuse him. As it has been said, whosoever washes his hands of his
354
H. W. BAILEY —
doruyi 6 mo yas ema ke az rOst votvuni ta. va ke hokemfl sevota
doruyi maslahat-omiz vaterun az rost votviiii i bad.
harki pOdeso 6 vekra ci o vevaja
heif-on ke ;^ub vevaja
bar toyi farldun navesta ba
hes vayt takya bar molki donvo maku
ke ;:^eili mardom mo kore seka v6 kuste boien cun vodem va;^ti
martdn voska va juni pok . . . ce bar ta^t but ce bar ;^ok.
bes fary nakre ma;)^sudos mOnegl vodem va^tl marf'unos ba ce
bar tajfti zarrin bud ce bar zevin.
IV
ovarde and ke enosirovuni odeko dar yak sekorl yak seid oska
va dar hemun suat seberiflnka nemak naba nokarOs se dg ngft to
nemak blora. enosirovun osv6d nemakro beyeimati manoseb vgrin
to rasmi nabut va dg rl be^arobi nasQt. sovOd az mo re ce yaloni
z6h6re but. sev6d asli zolm dar doni kemi beda. harki oma bar u
zrdd ba. To be mo yoyat ras6d.
ager ze bOye podisd yak sovi be^ra
nokarun dra;^t ro az zevin bar etoren
dar doni setemkor az bein esut
bar u lanatl mlnid to poyadOr.
V
gadQI podisOiro osdid adoros osvod ke ta ;;(eili pul dori va yda va
melki zl6d dori emmo me ki devari ta he heel nadore. davlati yaro
adem beme. podisO osvOd beyaki ke pfili sio osaden. gadu osvOd
ke mb ce ma'ni dora ceraki beme sahmi devarona omnatie. p5disb
yanda.s ka osvOd ki gap nakud ki age devardni be vefahmen mo dl
beta naresi. yo ba.
VI
yake haruja .sasta nfin seherit. yak rd yaki az refiyO adovos osparsud
ke haruja .sasta ndnro ce sevohri. adovds osvOd ke nuni beri xa taya
kre . . . va yak ndnl bero pete va diita ndni bero yarz ete. refiyos
osvdd aslen moltefet na boie ke ta cicidvod. vater beva ke multafet
YAZDI
355
own soul, whatever he has in heart, will say it. The king asked his
servant, What does he say ? One of them a minister who was present
said. Lord, he says that you have given a foolish command to kill
him. When he heard this, he pitied him and turned from shedding his
blood. Another minister who was hostile to the unfortunate said,
Do not the children of our kind know that in the presence of kings one
must not speak save the truth ? He abused the king and spoke ill
of him. The king frowned and said, I was more pleased by his false-
hood, than by your truth-speaking. And he who is a ruler has said that
a falsehood mixed with good is better than bad truth-speaking.
Whatever a king does or says,
Alas, if he speaks aught but good.
On the arch of Faridun was written ; —
Never trust the wealth of the world, for many men have laboured
and been slain. AVhen a man has reached the time of death and the
pure soul (has resolved to go), what matters it on a throne or on the
ground ?
It makes no difference to his purpose when a man has come to the
time of death whether it is upon a golden throne or upon the ground.
IV
It is related that Anusirvan the Just on a hunt secured his quarry
and at once he roasted it. There was no salt. He sent his servant to
the \Tillage to get salt. Anusirvan said. Buy the salt at a fair price
that it may not become customary and the \’illage fall to ruin ! They
said. From this course what ruin will come ? He said. At the beginning
the oppression in the world was little. Whosoever came increased it,
till it has come to this pass.
If a king eat an apple from a peasant's garden,
his servants will pull out the tree from the earth.
In the world the oppressor shall perish.
On him curses rest for ever.
V
A beggar saw a king. He said to him, You have much money, and
house and great property, but I who am your brother, have nothing.
Give me of your wealth. The king told someone to give him some small
coins. The beggar said. What does this mean ? WTiy do you not give
356
H. W. BAILEY
be. osvod 6 yaki nuni ki beri gos ekre va 6 yaki ki dir evene bexevari
miram ete va 6 diite nunro ki pete bebedar 6 merom ete beavaze
Oiki ie s5 meidoda dar kasogi va 6 diite ro ke yarz ete beporoguni
;(adoin ete ke dar vayti pin bedardi me va^ren ve avazo somaten.
yo ba.
Variants proposed by Mazkuri : —
beva : veva.
be : vebe.
[vefahme ki maysude cici ne : variant of Qobad.]
ete : ate.
VII
tojeriro va;;ftiki var-sekestaba bezur toman yarz osdost. adovof
p6r osvod ki mo gapero adovoi heski nava. poros osvod ki etdat
ekre va navajime besarti digi foide mo gafero ki nakude cici ne. bedaros
osvdd vaxtiki ya mosibati diita nabud yakis abamiyati nadora va
yaki bes zarari moia ne va dovomis semotati bemsdiegdn on.
sar —
hesva^t nava yemi ;(;aro adovoi dosmenbnot
ki henomgi az lopi ta bari nehomda ;(asuli kre
Variants (of Mazkuri) : —
nava : mava
foide mo : poida mo
sar : beit
VIII
yag ruj ardavfin ve savdrun sflane sekor. yaki gdr e ;(er tu viovQn
e devdrt. erdesir ;:(ad6 por e mase arda\iin sedumbol davOden. erdesir
yaki tiros kOste tu telos gi to parrose prQ so va otrafe bar omct va gur
e x^r sari yogas mart, ardavun ve savOrun sar rasoden. vaxtike mo
dost, o tir sodid heirun menan. ve soparsod ki ke m6-s kost. erdesir
osvod ki me mikarta. por I ardav'Qn osvod ki na me mikdrta. erdesire
yds oma ve be pori ardevun osvod ki sujoati mero xado druv ve tayallob
be x^ maben ve x^ra azlz nuiku. mone vioviin mas-on gQra x^r ji
pur-an. bri mazenna va emtehun ya buri be Ji kusim to malim vebud.
ardevun moskel os bo ve ezmos nadod ge erdesir sovor asbe bud. va
ardavdn erdesir osnbpte tu tavila. osvod ki hoderi asb bud ve bidi
meidune gu v5 capta 6 asp-davums naut oma.
YAZDI
357
me a brother’s share ? The king laughed and said, Be silent, lest if
your other brothers understand, this too should not reach you. Ended.
VI
A person bought each day sis loaves. One day a friend asked,
\Vhy do you buy six loaves every day ? The person said, I keep a loaf
for myself . . . and another one I give away, and two I lend. The
friend said, I have not fully understood what you have said. Say it
more clearly that I may understand. The person said, One loaf I keep
for myself and the one which I throw away I give to the sister of
my husband, and the two which I give, I giv'e to my father and my
mother in return for what they gave me in childhood and the two
that I lend, I give to my children that in the time of my old age they
make eat . . . and give it me in return. Ended.
VII
A merchant when he became bankrupt owed a thousand tumans.
To his son he said, Tell no one of it. The son said, I obey and will not
speak of it, on condition that you say what is the advantage of not
speaking of it. His father said. So long as a calamity does not become
two, it is not important, and one is the loss of capital and the second is
reproach of neighbours.
Verse.
Never tell your own trouble to your enemies.
Since from your mouth he will be rejoiced.
AHII
One day Ardavan went hunting with his horsemen. A wild ass
passed in the desert. Ardasir followed with the eldest son of Ardavan.
Ardasir shot an arrow into its body so that the feather went in and
came out the other side and the ass died on the spot. Ardavan and
his horsemen came up. When they saw the hand and the arrow they
were amazed, and he asked, Who shot this ? Ardasir said, I did. The
son of Ardavan said. No, I did. Ardasir was angered, and said to the
son of Ardavan, Do not take to yourself my prowess by falsehood and
deceit, and do not exalt yourself. Here the desert is large and there are
many other asses. As a test and proof, let us shoot a second time,
that it may become clear. Ardavan was angered, and refused to give
permission for Ardasir to ride a horse. Ardavan sent Ardasir to the
stables. He ordered that he should be a keeper of horses, and should
not again come to the racecourse for ball and stick and horse-racing.
358
H. W. BAILEY
IX
yaki rubas yaki dera^ti raz os did. sevlost ki raz ve;;^ra vall das
narasod. cem bur ]i arvast vail vipoida bo. oyari kor bidi hes koros
naka. osvod ki mo raze tr6s un. age ve;^re holi nadore be. pa vdterun kl
nayre.
X
yaki obed yak ruj yaki kalecca se mdye bedabo. nOkerhOs
vot ki ke sisdeda. hesta JavOp so nadod. obed osvOd ki me be bar
vaki yak! cflvi yak-yadd ete. harki dozz 6 cQvos yak kemogi massere
bud. benO bar in be bar yaki yaki cuvi bem-kad osdod. oige doz
bo yak kemogi sari cuvos se bar k«. ruje bad oige cQvos kasok bo
molim bo ki dozz o. yo ba.
XI
yaki merddge gadQ yiolos ka ki age bona vekra ki gong-on vester
pul se gir etud. benO bar in yaki taytas sided o ri-s os menes gong
6 se gall mol ven. yaki be merdogi gadQ gi d5smenos bo seviost gi
se rosvd vekra. yaki merdog otrape davdrt. yaki aldis d6de gong
6 gadQe ki dosmenos bo v6| se sar dod ki mo gong na ha. drQve vaja.
merdog i gong osvOd nd td dr live vajf. mone malim bo gi gong na
ha. rosvo bo. yo ba.
XII
yaki pOdeso dd\Ti yaki sdr 6s pnort. seviOst ki yr6v ekra. pei6m
os nepte tu .sdr ki kuri be yenun nadora. hdrci cume yeimati doren
asnen ve herdo bar sen. yenun Omoien 6 mire yd so poska 6 bar soen.
podeso svot gi cici do posdo.^ sovot cume yeimati mo podesos pdmod
ve yendds ka ve mie bamd osbaysOd. yo ba.
XIII
yogi yatarnok
yaki kastibun be rafiyois osvot ki me bdero mero bumaso memasom
bamd tu kesti mdrten. rafiyos osvot gi me be td nasihat ekre bes
bdr tu kesti masu bri gi yogi yatarnoki ne. kestibunos parsOd ki
badero mero memaso bumase td kojuL mdrten. osvod bamd tu davaji
mdrten. be ta nasihat ekre hes bdre tu davaj masu bri gi yogi
yatarnoki ne.
1 Read poska.
YAZDI
359
IX
A fox saw a grape vine. He wanted to eat the grapes, but he could
not reach them. He leaped up several times, but it was useless. At
last he did nothing more. He said, These grapes are soirr. If I ate them,
I should be ill. Then it is better for me not to eat them.
X
A religious man one day lost a jar. He said to his servants, ^\^lo has
taken it ? No one replied. The religious man said, I will give each one
a stick of the same size. WTioever is the thief, his stick will become a
little bigger. Then he gave each one a stick of the same size. The one
who was the thief cut off a small part of his stick. The next day the one
whose stick was smaller was evidently the thief. Ended.
XI
A beggar thought. If I claim to be dumb, more money will come
to me. So he took a hoard and wrote on it “ dumb ”, and put it around
his neck. Another beggar who was his enemy wished to show him
a cheat. A man passed by. He gave money to the dumb one. and the
beggar his enemy cried out. He is not dumb. He lies. The dumb man
said. No, you lie. Then it was evident that he was not dumb, but a
cheat. Ended.
XII
A king invested a city. He intended to lay it waste. He sent a
message into the city saying, I have no concern with women ; whatever
valuable thing they have, let them take and to-morrow go out. The
women came and carried (?) their husbands and went out. The
king said. What have you on your back(?). They said, valuable
things. The king understood and laughed and pardoned them all.
Ended.
XIII
A dangerous place
A ship's captain said to his friend. My father and mother, grand-
father and grandmother, all died on a ship. His friend said. Let me
advise you : never go on a ship, since it is a dangerous place. The
captain asked. Where did your father and mother, grandfather and
grandmother die ? He said. They all died in bed. Let me give you
advice : never go to bed, since it is a dangerous place.
H. W. BAILEY —
.S6()
XIV
yakl merdoge piri mosolvun resos se ra»gi slo Ikdrtebo vail
mldl saros sio [read : svid] bo. yak ruj yoptudabo. emoma se tep
kdft. centa rafiyos rasoden sodid ke midl saros sevid 6 vali midi resos
siOne. yeili taajjub so kd. merdogi pir so bldOr kd so pdrsbd cera
midi sarod svid on ve midi re.sod sione. JavOb osdod ki dar vaccegi
midi sar me bar! 6md« vali midi resom dar vis-solegi bari Omda bri
mone, gi midi sarom pirteri midi re.som o.
XV
yaki pdrog bo yiOlos ka gi yeili bosyur on bedveri osvOd gi me
yaki yuz dore td ]i yaki dori mene yaki ve dota seta bud. daveri
kasok osvod na dota }iiz dbriin. daveri mas osvod na seta \iiz dorim.
mi mmuine beder 56 sar rasod. pore kasugi coyoli e p6ri masos ka.
bederos v5t yaki iniz meyre, yaki Jl ta veyo sewom Ji In veyra gi
va]d seta ne.
XVI
yaki sarbozi pir dar jengi p6 se mdrti 6mdab6 ve avazos poie
cuvis dost [var. : dort]. yak rfl] s6 viost ke v^sute yadas dar mei’Bnl
ra motdr se ri barso ve p6ye cuvi se mdrtoma. voj 5e sar d6d ki ei
motorci me p6 nadore me tu motdr nenom ve bu. mutorci Ji gos
se gaf ka o 65n6de tu motdr 6 osbiirte tu marizyflna. va^tige dokt6r
osdid osvOd ke voje najjur kren ki pu-s vesiija. mOtorci yandas ka.
o5v6d ki me tabib mau na najjfir. yo ba.
XVII
Lalai for children of two years
gadu oma dari yOna
nfinom dud ya.sos 6ma
yados sa v6 sagos oma
coyom ka bados oma
For children of four to six vears
6 lalala nemflnot bud
memii korbfmi jiinot bud
babu benda yolumot bud
babu harunl numot bud
ki memu korbuni jOnot bud
babd harQni ^ numot bud
‘ harun is of uncertain meaning.
YAZDI
361
XIV
A man, an old Musulman, dyed his beard black, but the hair of his
head was white. One day he fell asleep, and his turban fell off. Several
friends came, and saw his hair was white, but his beard black. Thev
were surprised. They awakened the old man and asked, Why is the
hair of your head white, but the hair of your beard black ? He replied,
Tn childhood the hair of my head came, but the hair of mv beard came
at the age of twenty years ; therefore the hair of my head is older
than the hair of my beard.
XV
A boy thought he was very clever. He said to his brother, I have
one walnut, you too have one. Xow one and two are three. The younger
brother said, No, we ha\'e two walnuts. The elder brother said. No
we have three walnuts, ileauwhile their father came. The younger
boy complained of the elder boy. The father said, I will eat one walnut,
you too will eat one, the third he shall eat who says there are thiee.
XVI
,\n old soldier had a leg broken in war, and in place of it had a
wooden leg. One day he wished to go to his house. In the road a motor
went over him and his wooden leg was broken. He called out, Slotor-
man, I have no leg, put me into the motor. The motor-man agreed
and put him into the motor and took him to hospital. When the
doctor saw him he said, Call a carpenter to mend his leg. The motor"
man laughed, and said, I want a physician not a carpenter. Ended.
XVII
A beggar came to the door of the house.
I gave him bread, he was pleasecl.
He went and his dog came.
I drove him off, he was angry.
VOI.. vm. PARTS 2 AND 3.
24
Does Rhari Boli mean nothing more than Rustic
Speech ?
By T. Grahame Bailey
(Before proceeding to the discussion of the question I would draw attention to
the important quotations from Dr. J. B. Gilchrist on pp. 366, 7 below, which show
that at least four times in 1803, and twice in 1804. he used the name Khari Boll,
and tell us in what sense he used it.)
T his question arises out of some remarks made by Professor Abdul
Haq of tbe Osmaniya University, Hyderabad, Deccan, who,
criticizing views on Khari Boll ( = KB) which I had expressed in
ray Hist, of Urdu Lit., pp. 5, 8, 9, 13, said ; —
hame is se lAushi hul ki daktar sahab ne is mugalite ko rafa’ kia
hai, lekin asl galatl me yeh bhi mubtila hai : khari aur khari ka farq
inho ne bahut sahih bataea hai, lekin nia'ne taqriban vohl rakkhe
hai jo khari ke hai, ya‘ne murawaja, ‘am, mustanad (standard)
zaban ; aur dusra gazab kia hai ki Khari Boli ko ek lias zaban qarar
dia hai, aur us ki do shaldie batai hai, ek Hindi aur dusri Urdu . . .
Khari Boli ke ma'ne Hindustan me ‘am taur par gavari boli ke hai
jise Hindustan ka baeca bacca janta hai ; voh na koi lAas zaban hai,
aur na zaban ki koi shali. {Urdu, July, 1933, p. 590.)
“We are pleased to note that Dr. Bailey has corrected this mistake
(made by some Europeans, of confusing khari with khari, T.G.B.)^
but he too has fallen into what is essentially the same mistake ; for
though he has clearly shown the difference between khan and khari.
he has given khari almost the same meaning as khari, i.e. current,
common, accepted ; and he has made another amazing statement —
that KB. is the name of a particular language ; he has further divided
it into two branches, Hindi and Urdu. In Hindustan KB. usually
means ‘ rustic speech ’, a fact which every child in Hindustan knows.
It is not a particular language or branch of a language.”
I must stop here to correct the statement that I have given
khari and khari “ almost the same meaning ". I have never done
so. khari means “unadulterated” or “pure”, and while it may be
applied as an adjective to a language, it has never been the name of
any variety of speech, whether rustic or not. The word khari means
“ standing ”, and when first used of a language appears to have.
364
T. r;. BAILEY —
signified " current Only it must not be forgotten that it has
never been used of any language except that which we know as KB.
That the word does mean " standing ", and lias nothing to do with
khari “ pure ”, is further evidenced by the corresponding words in other
Hindi dialects or languages. I am indebted to you. sir (Sir George
Grierson) for a reference (in a private letter) to Kaintii Prasad Guru's
HinT) Vijdkaran. p. 25. We read there that ” in Bundelkhand KB.
is known as tharh boll This word tharh of course means " standing ".
A^ain, Dr. B. S. Pandit, whose native language is Marvarl, told me
that in Marvarl KB. is called “ tliath boll ”, where thath has the significa-
tion of “ standing ”. We thus have three names for this dialect,
and in each case it is called “ the standing language ”.
In Urdu for January, 1934, p. 158, Pandit Manohar Lai ZutshI
replies to Professor Abdul Haq, and says he is mistaken, for KB
undoubtedly is the name of a language. The Professor in a note on
p. 160, rejoins “ in my opinion KB means simply the opposite of
polished and literary ; it is used in that sense to-day, i.e. rustic speech.
Lallu Jl Lai probably used it with the same meaning. European writers
have fallen into error about it, sajdng it is a particular language.
The Hindi authors quoted by Paiidit ZutshI have merely followed
these Europeans ”.
It will be noticed that by the phrase “ in my opinion ” and the
word “ probably ” he has toned down his previous statements, but
ev'en so the matter rests simply on his assertion ; he gives no references
and quotes no authorities, nor does he name any of the Europeans
who supposedly have misled later generations of Hindi scholars. In
matters of Urdu his opinion commands re.spect, for Urdu is his mother
tongue, and he has devoted his life to Urdu scholarship ; this, howev'er.
is a question not of Urdu but of Hindi, and it must be decided from
a study of Hindi literature.
In Urdu literature the term has no meaning, for it does not occur ;
it has practically never been used in an Urdu book, nor is it found in
Urdu tnzldras (anthologies). Ev'cn Urdu dictionaries rarely contain it.
The Farhang i Asafiya, of which Urdu scholars speak with bated
breath, does not mention it. The meaning ” rustic speech " which
we are told every child in Hindustan knows, is not known to the
compiler of the voluminous Nur nl Lugdt. for all he says is Khnrl
Boll is conversation in the .style and pronunciation of men {mard,
men, as opposed to women ; T. G. B.) ; nor is it found in ‘Abdu'l
Majid’s huge Urdu dictionary, Jdmi' ul Lugdt, which explains KB
DOES KHAR! BOLI iFEAX XOTHIXO MORE THAX RUSTIC SPEECH 365
simply as mardo H hoTi, ‘‘the speech of men." We see then that
the compilers of the two large modern Urdu dictionaries, themselves
Indians, have never heard that meaning of KB which we are told
every child in Hindustan knows. There is nothing about rustic speech
in either.
In modern conversational Urdu usage Idian boll occasionally does
mean, not exactly village speech, but uncouth, boorish .speech,
though the dictionaries know nothing of this. But again we must
remark that Urdu usage does not concern us. We are dealing with
a Hindi term, and want to know what it signifies in Hindi. In my
History of Urdu Literature I gave the term its literary meaning, using
it exactly as Hindi writers do to-day.
Three points arise ; —
(i) WTio are the Europeans who have used the name KB ? And
in what way, if any, can it be said that they misled Hindi authors who
followed them ?
(ii) ^\^lat have Hindi writers in the last hundred years meant by
the name, what do they mean by it now, and what do they think
Sadal ilisr and Lallii Lai meant by it ?
(iii) ^^^lat did Sadal Misr and Lallu Lai, who were the first Indians
to use the term, mean by it ?
(i) The idea that certain Europeans have led Hindi writers astray
liy their statements about KB is strange. It would have been helpful
if Professor Abdul Haq had told us who they are. The fact is that
Europeans have rarely mentioned the name.
I have recentlv' made the very interesting discovery that Dr. John
Gilchrist used the term KB at least four times in 1803, the first
vear in which any Indian is known to have used it, and twice in the
year following. He therefore shares with Sadal Misr and Lallii Ji
the honour of priority'. In fact, as he wrote the name four times
in 1803, and they only once, he deserves it perhaps even more than
they.
What happened is clear. He was Profe.s.sor in the College of Fort
William for four years, and for nearly the whole of this time Lallii
and Sadal Misr worked with him. He learned the name from them,
and in his dailv intercourse with them had every opportunity of finding
out its exact meaning. He often spoke of Hindustani as the collocpiial
speech of India or the grand popular language of Hindustan. He said
on several occasions that it had various styles. The court or high
stvle was Urdu, full of Arabic and Persian. At the other fxtreme
366
T. G. BAILEY
was the “ pristine or rustic idiom of that extensive language indefinitely
called Bhasha ”, while between them came KB. He has told us
further that in order to facilitate the transition from Urdu to Bhasha
he had caused a KB version of Sakuntald to be prepared.
The state of affairs, as he saw it, was this. In the towns, especially
those with a large Muhammadan population, Urdu was the ordinary
spoken language, in the villages some variety of Bhasha, while KB
or even simple Hindustani, was the language which appealed to
Hindus, particularly those away from Muslim centres. KB, owing
to its avoidance of Arabic and Persian words was compelled to use
words derived from Sanskrit which were familiar to the rural popula-
tion. Gilchrist states that the desire to teach these words to his
students was one of the reasons for bringing out books in that dialect.
To this extent it has, as compared with Urdu, a rural appearance.
There does not, however, seem to be any evidence that in those
days the words kharl boll in themselves meant village talk. In no
books of that or any other period do we find such expressions as
“ the kharl talk of Bengal or Madras or the Panjab or of English
villages ” ; one does not find “ so and so has a khard pronunciation ”
or ■' his conversation is very kharl ”. Now if khard (fem. kharl)
meant simply gdvdrl, rustic, one ought to be able to say all these
things. The fact is that Hindi writers always used KB as the name
of a dialect, and Urdu writers never used it at all.
The testimony of Gilchrist's Bnglish-Hindastani Dictionary (1786,
2nd ed., 1810 ; revised 1825) is important. Under “ country ” he
has the entry “ the language of the country, opposed to the town
hahur kee holee ” ; under “ colloquial it has (1810 ed.) rozmurru,
moohavuru. In neither case is kharl boll given as a translation, nor
do we find it under words like rural, rustic, etc. So far as I know, it
does not occur anywhere in the dictionary or in any of the manv
vocabularies which Gilchrist prepared, though khard with the common
meaning of " standing " is frequent.
Similarly in the numerous English-Urdu or English-Hindi
dictionaries which have been publi.shed, one never finds rustic, rural,
or country speech translated by kharl boll.
As Gilchrist’s early references to KB are of great interest, I quote
them here : —
(1) The Hindee Story Teller, vol. ii, 1803, p. ii ; “ Many of those
(stories) are in the Khufee Bolee or the pure Hinduwee style of the
Hindoostanee, while some will be given in the Brij B,hasha.”
DOES KHAEI BOLI MEAN NOTHING MORE THAN RUSTIC SPEECH 367
(2) The Oriental Fabulist, 1803, p. v. : “I very much regret that
along with the Brij B,hasha, the Khuvee bolee was omitted since this
particular idiom or style of the Hindoostanee would have proved
highly useful to the students of that language.”
(3) ib. “ the real K,huree holee is distinguished by the general
observance of Hindoostanee Grammar and nearly a total exclusion
of Arabic & Persian.”
(4) ib., p. vii : (The learner) “ will find another specimen of the
K,huree bolee in the Story Teller, p. 24.”
(5) The Hindee-roman Ortkoepigraphic Ultimatum, 1804, p. 19
(foot) : “ Another version of Sukoontala in the K,huree Bolee, or
sterling tongue of India. This differs from the Hindoostanee merely
by excluding every Arabic & Persian word.”
(6) ib., p. 20 (foot), 21 (top) : “ The Prem Sagur, a very enter-
taining book, rendered with elegance and fidelity from the Bruj
B,hasha into the K,huree Bolee by Lalloo Jee Lai expressly to effect
the grand object of teaching our scholars the Hindoostanee in its
most extended sense, and with proper advantages among the grand
Hindoo mass of the people at large in British India.”
Gilchrist always marked in one way or another the cerebral r
which occurs in the name Khari Boll.
In 1814, Lieut. 'William Price published a K,huree Bolee and
English Vocabulary of all the principal words occurring in the Prem
Sagar ” of which the Directors remarked “ these (words) are in constant
use in other K,huree Bolee and Bhakha compositions ”. Although
the name KB occurs in the Introduction to the Prem Sugar, it is not
given in the vocabulary. The only meaning given to bhari is chalk,
a signification, which, so far as I remember, is not to be found in the
Prem Sugar.
This vocabulary was reprinted in Hindoostanee Selections, 1827,
2nd ed. 1830.
The next whom one should quote is Garcin de Tassy. In his
Hist, de la Litt. Hindoiiie et Hindonstanie, 1st ed., vol. i, p. 307, he says
that Lallu’s Prem Sagar was “ non pas en urdu, mais en khari-boli ou
thenth, c’cst-a-dire en hindoustani pur, en hindoiistani hindou de
Dehli et Agra, sans melange de mots arabes ni persans. This is a
paraphrase of Lallu’s own words, but, mistaking kliarl for khari, he
interprets it of LalhVs phrase ” omitting Arabic and Persian words ”,
thinking that it means “ pure language ”. G. de T. does not mention
Khari Boll at all, but speaks of khari, “ pure.” i.e. without mlecch
368
T. G. BAILEY
“ unclean ”, words of non-Sanskritic origin. He WTOte the words quoted
(and almost the same words on p. 1 of the Introduction) in 1839, and
repeated them in 1870 ; as they were French, not English, the con-
fusion between Jchafl and Marl passed unnoticed in India.
Eastwick, in his vocabulary, 1851, says that Marl boll means
Tcliarl boll “ pure language ”.
Platts, Urdu. Diet., 1881, under khard has “ kharl boll, \’ulgar
kharl boll, pure language ”.
The language which Hindi authors call KB English writers prefer
to call High Hindi or Classical Hindi, names which correspond to
nothing in Hindi itself.
(ii) The name KB is Hindi ; the first Indians to use it were, as we
shall see below, Lallu Ji Lai in 1803 and 1818, and Sadal Misr in 1803 ;
it is in constant use now by Hindi writers. We are therefore bormd
to ascertain what they mean by it. The Urdu meaning, if any, does
not matter.
Have any of them given it the sense of rustic speech ? If so, when
and where ? For many years after the time of Lallu and Sadal Misr
they did not employ it at all. The first I know of to use it since those
days was Raja Siv Prasad in his Introduction to 1867.
He regarded it as essentially artificial and literary; in fact, he
says that LaUu Ji, though he strove to preserve its literary character,
yet failed sometimes to exclude the Braj village words to which he
was accustomed in his own speech. His words are ; “ Whether this
new dialect, the Prakrit enriched with Persian and Arabic words, be
called Hindi or Hindustani. Bhakha, or Brai Bhakha, Rekhta or
Khari Boli, Urdu or Lrdu-i-Muaalla. its seeds were sown here by the
followers of JIahmud of Ghuznee ” (op. cit., p. vi).
On p. xi he goes on : When Dr. Gilchrist asked Mir Amman and
Lallu ji Lai Kavi, to write some vernacular books in prose, they must
have felt themselves very puzzled : it was quite a new thing to them.
They wrote, but they both WTote in an artificial language.” .Six
pages further on, p. 17, he says : “ Lalliiji has not allowed foreign
words, Persian or Arabic, a place in his book (Prem Sdcjrir. T. G. B.),
but he could not so well keep clear of the patois of his native place
Agra.” He has the same statement, but in Hindi, on p. 32, of Part I
of his yai/a Gutkd, 1900 cd., finst published 1867, *• he wrote in the
khari boli of Agra ; although he excluded Persian and ^Vrabic words,
he was not able to keep out Asra village words.”
By Hindi writers the name KB is given to a particular language
DOES KHAEI BOLI MEA^J NOTHING MORE THAN RUSTIC SPEECH 369
or dialect, viz. that form of Hindi which is used in every-day Hindi
prose (and increasingly in verse), the Hindi which we find in all Hin di
magazines, in translations such as the Hindi Bible, scientific works
and all school books. This fact is so well known that proof is hardlv
necessary. In an article {JRAS., Oct., 1926, pp. 717-723) I mentioned
and quoted twelve Hindi authors to this effect. This is the ordinarv
meaning of KB, but the Urdu language itself is sometimes spoken of
as a branch of it. KB is contra.sted with Braj, AvadhT, and other
Hindi dialects.
There is no need to labour this point ; I will content myself with
one or two further quotations, to illustrate what they mean by KB,
and to show that they do not think of it as rustic.
The IMisr Brothers in Misr Bandhu Vinod, vol. i, p. 119, say that
“ Sital (a.d. 1723) wrote all his poetry in KB Sital's language is
far removed from rustic speech.
Badri Nath Bhatt in Hindi, p. 31, after mentioning that he lives
within twenty yards of Lallu’s old home in Agra, says that every Hindu
household in Agra city speaks the same language as Lallu's in Prem
Sugar, the only difference being that which naturally exists between
literary language such as Lallu's, and conversational speech, such as
is heard in the Hindu homes. He calls Lallu's KB literary^, not rustic.
The best known of modern Indian writers on Hindi literature and
languages, Shyam Sundar Das, says in Hindi BJtdsd hi Vikas, p. 54,
“ between 1250 and 1450 a.d. the older Hindi dialects gradually
assumed the form of Braj, Avadht, and KB.'’ and on p. 55, “ KB
was used for poetry not only by Musalmans but by Hindus also."
Hama Sankar Prasad in Hindi Snliilga kd Sanksijil Itihas. p. 128.
writes “ Sadal Misr and Lallu wrote in KB mi.xed with Braj bhasa ".
He thus contrasts Braj and KB as two distinct dialects.
There is an important reference in Eama Kant Tripathl's Hindi
Gadija Mhndsd, p. 33 of Introduction, '' the language of the Prem
Sugar is adorned to this extent that all through it there is the splendour
of Braj bhasa. Not only so, but it is characterized by a great pomp of
words and by poetical style ; it is not the plain idiomatic language
of conversation, but poetical prose."
From these c|uotation.s and from those in the article referred to
(JRAS., Oct., 1926), and indeed from the works of any Hindi author
who writes on the literature, it i.s plain that KB is regarded not as
rustic speech, but as a dialect of Hindi, and practically all Hindi writers
would denv Siv Prasad s statement that it was artificial.
370
T. G. BAILEY —
(iii) We come now to the important question : What did Sadal
ATisr and Lallu Jl mean wlien they said in 1803 that they were
translating into KB ? Did they mean “ into rustic speech ” ?
Sadal Misr in the Introduction to his XasiketopakJiyan says ; “ Some
people cannot understand the Nasiketopakhyan because of its being
in Sanskrit, so I have translated it into KB."’
Let us recall the facts. Lallu belonged to Agra, Sadal ilisr to
Arrah (Ara) in Bihar, 450 miles away. To get from the former to the
latter we must leave the Braj area where Agra is situated, pass through
the country where Bundeli and KanaujI are spoken, into the AvadhI
country, and finally after entering Bihar traverse the Bhojpurl region
to a few miles west of Patna, the capital. The only rustic speech
Sadal Misr knew was that of his native Arrah and the country round
it ; it was entirely different from that of Agra ; the former was Biharl,
the latter Braj, and the whole country of still another language,
AvadhI, lay between. Ram Candr Sukl in his Hitidi Bhdsd aur Sdhiti/a
(at the end of the Sabd Sagar. p. 210, also published separately) tells
us that KB in those days and previously was the language of educated
and polite conversation among Hindus from Delhi to Bihar. It is
interesting to note that Sadal Misr, though he lived so far from the
real home of KB, ^Tote it better than Lallu who lived very near it.
Lallu’s is too much tinged with his native Braj. The style of both
men, though simple, was literary, not rustic.
A dozen or so years earlier Sadasukh Lai, of Delhi, a man learned
in Arabic, Persian. Urdu, and Hindi, wrote KB still better than
Sadal Misr and Lallu. He wrote just the straightforward Hindi which
he was accustomed to talk to his educated Hindu friends, at least on
formal occasions.
We turn now specially to Lallu Ji. In the introduction to his
Prem Sngar he stated that avoiding Arabic and Persian words, he had
told the story DilVi Agre kl KB we, in the KB of Delhi and Agra.
Did he mean " rustic speech '• ? The rustic .speech of the district
round the two towns was different. The people round Delhi spoke
what we now call KB, tho.se round Agra, 120 miles away, spoke Braj.
If he had written in the rustic speech of the former it would not have
been the rustic speech of the latter. Secondly, like Sadal Misr, he is
literary not rustic. It is true that he .sometimes failed to e.wlude
Braj wmrds and forms (perhaps he never tried), but Braj forms are
not KB; indeed, we have seen above that KB is contrasted with Braj
Fifteen years after the Prem Sagar Lallu Jl wrote the Ldl Candrikd
DOES KHARI BOLI MEAX NOTHING MORE THAN RUSTIC SPEECH 371
a commentary on Biharl’s Satsai ; this was also in KB, and showed
less Braj influence. In the Introduction he distinguishes three dialects
in which he had written books, viz. Braj, KB, and Rekhte kl Boli
(i.e. Urdu). In his Braj and KB books he usually endeavoured to
avoid Persian and Arabic words, but in the Introduction just mentioned
he used them rather freely
Conclusion
We may sum up by saying ; —
(i) KB is a Hindi term, and its meaning must be sought in Hindi
writings.
(ii) By KB Hindi authors always mean a dialect, often, though
not always, a highly polished and literary dialect.
(iii) It is difficult to believe that KB means rustic talk, for if it
did it could be used of the village talk of any part of the world, and
one never hears this meaning given to it.
(iv) There is no proof that any European writer has misled Hindi
authors as to the meaning of KB.
(v) There is a little evidence that in conversational Urdu KB
sometimes means boorish and possibly uneducated speech, but
this is not supported liy Urdu dictionaries whether compiled by
Indians or by Europeans.
(vi) The early KB wniters, Sadasukh. Lallu Ji, and Sadal Misr
did not write in a rustic style.
Modern Literary Hindi
By A. Baraxxikov
nPHERE was a prevailing opinion in English linguistic literature
that the modern prose Hindi. High Hindi, or Kharl boli, was
invented by the English.
This point of \dew was maintained some time ago by such a highly
authoritative scholar as Sir G. A. Grierson. Thus, in the preface to
his work on the history of Hindi literature he says : “ The first half
of the nineteenth century . . . was the period of the birth of the
Hindi language, invented by the English, and first used as a vehicle
of literary prose composition in 1803 under Gilchrist's tuition, by
Lallu Ji Lai, the author of the Prem Sagar."
In another place of the above-mentioned work Sir G. A. Grierson
explains in what sense the term invention ” he uses should be
understood. He says ^ : “In 1803, under Gilchrist's tuition, Lallu
Ji Lai wrote the Prem Sagar in the mixed Urdu language . . . with this
peculiarity that he used only nouns and particles of Indian, instead of
those of Arabic or Persian origin. The result was practically a newly
invented speech ; for though the grammar was the same as that of
the prototj-pe, the vocabulary was almost entirely changed. This
new language, called by the Europeans Hindi, has been adopted all
over Hindustan as the lingna franca of Hindus, for a want existed
which it fulfilled. It has become the recognized medium of literary
prose throughout Northern India.”
Further investigations into the domains of the history of Indian
languages, investigations in which Sir G. A. Grierson himself has
taken such an eminent part, have shown that the history of the Hindi
and Urdu languages is much more complicated than it was supposed
even at the end of the last century. In conformity with the results
obtained by these investigations, Sir G. A. Grierson to a great extent
modifies his point of view with regard to the formation of the High
Hindi. In his article " Indo-Aryan Vernaculars ”, he says ^ : “ The
‘ The Modern Vernacular Literature of Hindustan, Calcutta, 1889, Introduction,
}i. xxii.
- Op. cit., chapter x, p. 107.
“ “ Indo-Aryan Vernaculars,” Bulletin of the School of Oriental Studies, 1918,
vol. i, pt. 2, p. 52.
374
A. BARAXXIKOV —
present form of literary Hindi or High Hindi, is a reversion to the type
of the non-Persianized vernacular of the Upper Doab, brought into
use by the teachers at the College of Fort William in Calcutta in the
early years of the nineteenth century. It was desired to create a
Hindustani for the use of Hindus and this was recreated by taking
Urdu, the only form then known, as the basis. . . . Owing to the
popularity of the Prem Sagar of Lallu Ji Lai, one of the first books
written in this newly devised speech, and also owing to its supplying
the need for a lingua franca which could be used by the strictest
Hindus . . . etc.”
A whole number of European authors recur to the version of the
English having invented a new language. However, after the
appearance of the above-mentioned article of Sir G. A. Grierson it is
generally said with reference to the invention of this language that it
was not invented by the English themselves, but by the teachers at
the College of Fort William, under the direction of the English. In
the Encyclojpsedia Britannica we read : “ the Hindi form of Hindustani
was invented simultaneously with Urdu prose by the teachers of Fort
William . . .” etc.
This European point of view up to recent times was popular
in India not only among Mussulmans, but among Hindus as well.
As an illustration one may bring forward the views of the brothers
Misra upon this question and suggested by them in their history of
literature of the Hindi. The brothers Misra say ^ ; Yarttaman gadya
he janmaddtd Sadal Misra aur Lalluji Lai nidne jdte fiai “ The parents
of modern prose are considered to be Sadal Misra and Lallu Ji Lai.”
Of course the brothers Misra are not considered to be the best authorities
in the domain of linguistic problems but we quote their opinion because
these words reflect the point of \dew widely spread in India itself.
However, after the war, when national and confessional relations
and contradictions became strained, this conception regarding the
origin of modern literary Hindi underwent a revisal on behalf of the
Hindu scholars.
It should be said that this criticism was not altogether fruitless,
as owing to it, it was possible to ascertain a series of historical facts,
which were heretofore unknown to science.
On the other hand one must acknowledge the fact that criticism
* Mhirnhandhui-inod athaid Hindi sahilya kd itikax, dvitlya bar, dvitlya bhag,
Lakhnau, sam" 1984, p. 8.52.
MODERN LITERARY HINDI
375
was and is carried on in a mode quite different to an unprejudiced
study of historical facts. Very often a series of circumstances are based
not upon a thorough study of historical facts, hut merely upon personal
impressions and emotions evoked by an upheaved national and
confessional proper pride.
It is impossible to fully envisage the question of the origin of the
modern literary Hindi in the frames of a small note, therefore I shall
only stay for the chief statements of Babfi Syam Sundar Das, one of
the most eminent connoisseurs of Hindi, the chief editor of the large
Hindi dictionary, Hindi Sabdasdgar, author of the first course of
general linguistics (bhdm-rijnan) in Hindi and other important
scientific works ; and also upon the views of the collaborator of
Syam Sundar Das in reference to the mentioned dictionary, Ram-
candra Sukla, expressed by him in a short essay on the history of
Hindi literature supplemented to the above-mentioned dictionary,
and Sri Jagannath Prasad Sarma — author of an essay regarding the
development of Hindi. All the three authors quite agree as to the
question of the origin of prose Hindi.
Their chief objections with regard to this question are as follows : —
In the history of the Hindi language neither prose nor the dialect
Khari boli upon which the modern prose literary language is based
are considered to be new phenomena. Hindi prose existed even
before the nineteenth century. As to the Khari boli dialect, its history
is ancient and has been known since the thirteenth century. Khari
boli was not invented by the Mussulmans, but it is a colloquial language
of the educated Hindu merchants, scholars, etc. ; the Mussulmans
merit the wide spreading of the language only. The literary language
of the Mussulmans, the Urdu, based upon the Khari boh, is only a
dialect of the Hindi language. The Urdu is merely an artificial language
and therefore cannot be considered as a basis of literary Hindi. The
honour of “ invention ” or introducing and spreading the prose
Hindi does not belong only to Lallu Ji Lai and to Sadal Misra who
acted under the guidance of the English, but much more so to the
authors that worked independently of them and whose language,
by the way, was much better than the language of Lallu Ji Lai, for
which reason this latter cannot be considered the inventor of the
modern prose literary language.
The above-mentioned Hindu scholars, as well as a number of others
who have written before and after them, point out that prose in the
Hindi language, carrying its incessant tradition since the nineteenth
376
A. BARAXXIKOV' —
century, is not considered to be an absolutely new appearance in the
Hindi literature.
It is true, that tbe works written in the literary dialects of Braj
and AvadhI in their majority are composed in verse ; but equally
with these, prose works are also known. Thus Sri Jagannath Prasad
Sarma after the historians of Hindi literature, points out ^ that one
of the works attributed to Gorakhnath, an author of the fourteenth
century, although, probably, it was written by one of his fol-
lowers, represents the earliest sample of Braj prose. In the middle
of the sixteenth century, in the same prose language were written the
Vartas of Vitthalnath as well as comments of some of his followers.^
However Jagannath Prasad Sarma himself points out that this prose
in Braj did not continue any further. The prose of a few commentators
— Vitthalnath's followers — is to such an extent formless and helpless
that it rather obscures the sense of the original than serves to explain
it. The helplessness and uncertainty of this prose was one of the causes
why the Braj prose tradition was soon smothered, not to be renewed
in future.
Still less successful was the attempt to create prose in the Kharl
boll, i.e. the dialect which served as a basis later on (in the nineteenth
century) to the development of the modern literary Hindi. The poet
Gang, who belonged to the court of the emperor Akbar, wrote a
little story in prose Kharl boll. This attempt of Gang’s was not upheld
by the following authors and remained a solitary instance.
For this reason Syam Sundar Das is quite correct in his remark,®
when he says the prose which existed in Kharl boll before the nineteenth
century was but nominal inatra).
The cau.'ie of the failure of these attempts to create a prose language
in Braj and Kharl boll are quite natural ; authors of the Hindi literature
of that period belonged in their majority to the high castes, whose
prose language was the Sanskrit, therefore there was no necessity of
creating another prose language which neither possessed such a
standard form nor the possibilities of being as widely spread as Sanskrit.
With regard to the question of prose the Hindu scholars themselves
come to the conclusion that single sporadic attempts of writing in
^ Hindi hi (jndya ioill ka likfis: Xagaripracarini Patrika. navln saiiiskaran.
bh;!" 11, ank 2, Sam' 1087, p. 187.
“ See also F. Y. Keay, A History of Hindi Literaturey pp. 30, 100, etc.
^ Adhunik Hindi gndya m// acary'z, Xagaripracarini Patrika. navui samskaran,
bhag 6, ank 1, p. 13.
MODERN LITERARY HINDI
377
Hindi prose, being of an interest by themselves and from the point
of view of the history of this literary language, were of no practical
consequence and could, in no manner, influence the future formation
of quite a new literary language, new in principle to the Hindi
hterature.
The circumstances of the Khari boll dialect are much more
complicated. The Hindus declare the question of its development to
be involved with many delusions, and its history is generally represented
erroneously. Such a point of view is expressed by Syam Simdar Das.^
This idea is repeated in a series of his works, in particular in an essay
on the history of the language attached to the large Hindi dictionary
Hindi- Sabdasdgar.^ The same idea is repeated by other scholars,
for instance Ramacandra Sukla in his essay on the history of Hindi
literature, supplementing the same dictionary.
One must acknowledge that in reference to the question of the rise
of Khari boli the Hindus are quite correct in many points.
Firstly, among the Indian Muslims, up to very recent times, an
opinion was widely spread that the Urdu had risen from the mixing
up of different languages and dialects, the speakers of which thronged
the court of the Mogul emperors. This point of view upon the Urdu,
mentioned in a well-known work of Mir Amman, was accepted by
several European scientists as well, part of whom up to the present
consider the Urdu to be an artificial language — originating from the
mixture of various dialects and languages.
Secondly, with regard to the formation of the Khari boli the
representatives of this opinion declare the literary Hindi to have
originated from the Urdu by the way of a mechanical exclusion of
Persian and Arabic words and by replacing them with Sanskrit,
pure tatsamas or tadbhavas.
Thirdly, up to very recent times a view was widely spread among
the Hindus themselves to the effect that the modem Hindi originated
from the Braj and was reformed under the influence of the Mussulmans.
Ramacandra Sukla points out that such opinions were expressed not
so long ago (in 1928) by the President of the Society Hindi Sahitya
Sammelan.
In order to show the error of these ideas, the Hindu scholars
quite justly point out that Khari boli was not a new language. In
^ Bkasa-i'ijuan, Kasi, sam° 1982, p. 342.
2 Hindi bhdsd kd vikds, p. 38.
PARTS 2 AND 3.
VOL. VTII.
378
A. BARANNIKOV
particular %ain Sundar Das says ' : “ Khari boli has existed from
the same time as Avadhi and Braj. The only distinction between
them is that the literary production both in Braj-bhasa and
Avadhi began to develop a long time ago. whereas that in the Khari
boli has begun quite recently.’’
Syam Sundar Das himself as well as Ramacandra Sukla and
Jagannath Prasad Sarma point out that, besides being used as a
spoken language, the Khari boli crept into literature from ancient
times.
Thus, Amir Khusro (1256-1325), known more as a Persian poet,
has written a series of small productions, riddles, etc., in the Khari
boli. And not only historians of the Hindi language and literature,
but also historians of the Urdu literature consider Amir Khusro’s
verses nearer to the Hindi than to the Urdu. Baburam Saksena ‘‘■
in particular says : “ These verses though they employ Hindi
words are scanned according to Sanskrit prosody and can scarcely be
regarded as UrdQ verses, though Persian words are found there and
here.”
After Amir Khusro the traces of Khari boli are partly observed
in the works of Kabir, Nanak, Dadu, in Bhusan’s “ Siva Bhavani ”,
in one of Lalit Kisori’s verses, in the verses of Sital Kavi, in one of the
already mentioned prose works of Gang — who belonged to the court of
emperor Akbar — also in Jatmal’s (about 1624) “ Gora badal ki
katha ” etc., and in other poetical works.
The above-mentioned authors give short extracts from the works
of the enumerated poets. These extracts, in a most convincing manner,
certify to the existence and development of Khari boli from the
thirteenth century ; thus, we may speak of the Khari boli as one of
the dialects of Hindi literature which was used in the latter to a very
limited extent and only in single cases as we may observe it in Amir
Khusro and in the works of the poet Gang — it was used as an
independent dialect and not as a special stylistic means as it appears
in Bhusan's works and in those of a few other Hindu authors.
The fact of using Khari boli in the Hindi and Urdu literature
undoubtedly testifies to the Khari boli being employed also as a
spoken language.
In this manner we may consider the antiquity of Khari boli as
^ Bhdsa-vijhan^ p, 342.
^ A History of Urdu Literature, Allahabad, 1927, p. 10.
MODERN LITERARY HINDI
379
proved. Therefore the idea of the Khari boll being a language which
has risen from the intermingling of different dialects, or having been
founded on the Braj, completely falls away.
The Khari boli is an independent idiom, having risen on the basis
of one of the local dialects. But whether it is founded on one of the
dialects used near Delhi, Agra, or Meerut, as is presumed by the
Hindu scholars, or whether upon the basis of one of the Panjabi
dialects, as is presumed by Grahame Bailey,^ does not enter the scope
of the present note.
However, the Hindu scholars, when looking into the correlation
between Khari boli, the modern literary Hindi, Urdu, and Hindustani,
come to conclusions with which one may undoubtedly disagree.
For instance, Syam Sundar Das, on the one hand, says ^ : “ There
are three forms of Khari boli : (1) the pure Hindi, which is the literary
language of the Hindus ; (2) Urdu, which is used specially among
Mussulmans and is their literary language as well as the spoken
language outside their homes, the language of educated Muslims and
several Hindus ; and (3) Hindustani, in which are used, without
any difference, words of both Hindi and Urdu languages and which is
used by all as a spoken language.”
In the above cited quotation %am Sundar Das considers Khari boli
as a more general conception and in the literary Hindi, Urdu,
Hindustani, he observes different aspects and different forms of the
language based upon the Khari boli.
Whereas, a little further on, in the same work ^ Syam Sundar Das,
following Pandit Candra-Guleri, states ; Urdu kdi bhdsd nahi hai,
Hindi ki vibhdsd hai “ Urdu is not a language, but a dialect of the
Hindi language ”.
Evidently Syam Sundar Das put a mark of equality between
Khari boli and Hindi as, if he had understood the term “ Hindi ”
as the whole of all dialects (from Bihar to the Panjab), he would have
had to acknowledge this dialect to be not only Urdu but the literary
Hindi and the Khari boli itself.
How to reconcile these contradictory statements is not clear.
E.xactly from this point of view Urdu is represented by the
scholars mentioned as a purely artificial dialect {krtrim vibhdsd).
Hindii scholars declare the merit of spreading Khari boli
^ “ Urdu, The Name and the Language,” JHAS., 1930, April, pp. 391-400-
2 Bham-vijnan^ p. 342.
® Bhasa-vijnanj pp. 346-7.
380
A. BARANNIKOV —
all over Northern India and over the Deccan to belong to the Muslims
as well as the fact of the Muslims being the first to use Eiharl boll
as a literary form. According to their opinion, however, Khari boli
in the hands of the Muslims obtained an artificial form ; owing to the
introduction of a large number of foreign, Persian, and Arabic
elements, the Urdu lost its national Indian character, all the more so
because the Urdu had taken up some of the elements of Persian
grammar.
One cannot deny the fact that the Urdu in the hands of the Muslims
underwent many changes, greatly removing it from the spoken
language. Nevertheless, these changes were quite natural, in so far
as this literary form began to be used for expressing ideas of another
Muslim culture, the scope of conceptions of which differ from the
notions peculiar to Hinduism. Besides this the Khari boli dialect,
modified by the Muslims and changed into Urdu, appears to be no
more artificial than the very same Khari boli in the hands of some
Hindu writers, who sometimes exclude the elements of tadbhavas,
replacing them by elements of Sanskrit tatsamas and by heaping up
most complicated compound nouns, some of which are composed of
ten or even more components. Both the overloading with Muslim
elements as well as with Sanskrit elements, especially with compounds
not properly pertaining to the spirit of Khari boli, attributes to it
an artificial form. The Hindus are justified by the Sanskrit elements
being national elements, but for the Muslims, educated on Arabic and
Persian culture, borrowing from Arabic and Persian languages it
also seems natural to have “ the national elements ” presented.
Hindu scholars insist upon the artificial, hybrid character of the
Urdu, especially so because very many authors interpret the expression
of Lallu Jl Lai : Ydvani bkdsd chor . . . khari boll me kah “ excluding
Muslim elements and narrating in a pure language ”, in such a manner,
that the modern literary Hindi is created by Lallu Jl Lai from Urdu
through excluding from it Muslim elements.
Trying to refute this statement, the Hindu scholars attempt to
prove first of all that the Khari boll existed separately, independently
of the Muslims and their “ artificial ” literary language ; secondlv —
that Lallu Jl Lai, who was employed by the English, was not the
creator of the modern literary Hindi.
It is quite natural and absolutely comprehensible w'hy the
Hindus objected to a simplified, purely mechanical, under-
standing of the formation of a modern literary Hindi, as though
MODERN LITERARY HINDI
381
it were formed exclusively by the substitution of some lexical
elements to others.
Even Lallu Ji Lai personally, when creating his work, does not
mention the fact of his departing from the Urdu, he says he writes
in a “ pure Delhi and Agra language ”, i.e. his point of departure was
that of the spoken language of these cities.
The Hindu scholars, however, try to prove that this pure Khari
boll language was mainly cultivated amidst purely Hindu
surroundings ; this language being used by Hindu pandits, sadhus.
merchants, etc.
These statements, of course, surmised in the ardour of Hindu
patriotism, lead the Hindu scholars to contradict themselves,
as they are obliged to acknowledge that the most ancient samples of
Khari boll are to be found either in the works of Muslim authors
(Amir Khusro, Kabir) or in those parts of the works of the Hindu
authors, where Muslims are represented.
Thus ^yam Sundar Das writes ^ : Hindu kaviyo ne bhi ajmi
kavitd me is khari boll kd jrrayog kiyd hai. Prdyah Musalmdnd kt
bdtcit we khan boll me likhte the “ Hindi poets also used the Khari boli
in their poetry. In general they used to render the conversation of
Muslims in Khari boli
Ramcandra ^ukla as well, in the above-mentioned essay on the
development of literary Hindi,^ says that from the point of view
of Hindu poets the Khari boll is understood to be specially a Musbm
language. Is se Bhiisan, Sudan ddi kaviyo ne Musalmdnl durbdro
ke jrrasang me yd Musalmdn pdtro ke bhdsan me is boll kd vyavahdr
kiyd hai “ For this reason Bhusan, Sudan, and other poets when
representing Muslim durbars or depicting the language of Muslim
personages used to employ this language ”.
It is quite evident the Hindu poets considered the Khari boli as
a special spoken language of the Indian Muslims. Undoubtedly
the Khari boli was the language of educated Hindus, but in the midst
of the Hindus it was a spoken language used by them (as is mentioned
by pandit Guleri and other authors), outside the home, for the spoken
home language was more or less coloured by local dialectical
peculiarities varying in different provinces.
Although the Khari boli has for long been used by Muslims
not only as a spoken language, but also in Muslim poetry, where it
1 Bhasd-vijnan, p. 343.
' Hindi sahitya ka vikds, p. 207.
382
A. BARANNIKOV —
appeared mostly in a form saturated with Persian and Arabic elements,
the Hindus absolutely ignore the use of Kharl holJ elements in the
Urdu hterature, although elements of Sanskrit tadbhava occupy quite
an eminent place with several authors of the Urdu literature.
Completely ignoring the use of Kharl boll elements in the Indo-
Mushm literature, the Hiudus speak of its existence only in the form
of a spoken language.
However, from the history of no matter whatever language, we
know if it lacks a literary language its spoken form is void of
stabilization both from a lexical point as well as in respect of grammar,
which in no case may be asserted with regard to the Kharl boli, for
it comes forward in quite a uniform style with various Hindu poets
who sporadically use it in their poems written in the Braj. Besides,
having no uniformity in the spoken language, Kharl boli could not
appear in such an analogous hterary form at the beginning of the
nineteenth century with authors who wrote in different parts of the
country, as did Lallu Jl Lai, Lai Sadasukh, and Insha Alla Khan.
Up to the nineteenth century the only form of a literary language
which to a certain extent directed the free development of colloquial
Kharl boll and communicated a certain steadiness and stabUity to the
latter, was the literary language of the Indian Muslims — the Urdu.
That is its historical merit wliich the Hindu scholars cannot efface.
Our acknowledging the exclusive role of the Urdu as a hterary
language, which has, during several centuries, influenced the Khari boli
spoken language and added to it a. certain stability, preventing it
from splitting up into a number of dialects, does not mean to assert
the hterary Hindi to he considered as having risen from the Urdu.
Both the Urdu and the hterary Hindi are grounded upon the spoken
Khari boli. The difference is merely that the Urdu began to develop
much earher, therefore it was able to exercise a strong influence on
the development of spoken Khari boli, attributing a certain steadiness
to it.
The undoubted influence of Urdu upon the formation of hterary
Hindi is also proved by the fact that Lallu Ji Lai, Munshi Sadal
Misra, Sadasukh and Insha Alla Khan all had a perfect knowledge of
the Urdu ; the priority in that respect belonging naturally to Insha
Aha Khan and Sadasukh, the former being one of the most eminent
poets of the Urdu hterature, the second the author of several books in
the Urdu and Persian languages.
It is quite comprehensible, therefore, that the Urdu did not
MODERN LITERARY HINDI
383
immediately affect the literary Hindi, but tbrough the medium of a
spoken language which, in the hands of the above-mentioned authors,
was moulded into a completeness of form only owing to their perfect
knowledge of the Urdu.
The patriotism of the Hindu scholars is manifested not only in their
inclination to deny an obvious fact of the effects of Muslim Urdu
upon the formation of literary Hindi, but also in that they try to
disparage the role of Sadal Misra and especially Lallu Ji Lai — the
authors who worked over the creation of a Literary Hindi under the
direction of the English, chiefly under John Gilchrist.
In the introduction to “ Prem Sagar ” of Lallu Ji Lai (edition
issued by Nagari Pracarini Sabha) the editor says that Lallu Ji Lai
is considered to be the first author of the Hindi prose as well as the
first writer in the Hindi in its modern form,” ^ Syam Sundar Das
and other authors repeatedly deny this role of Lallu Ji Lai. In this
manner Syam Sundar Das, in his course of general linguistics,^ says :
Lallujl Lai Hindi gadya ke janmadatd mane jale hai. Vdstav me unho
ne Hindi gadya ko ddhunik rup nahi diyd “ Lallu Ji Lai is considered
to be the parent of prose Hindi. In reality it is not he who
gave the prose Hindi its modern form.” The same is repeated by
%am Sundar Das in an essay treating of the development of the
Hindi, supplement to the dictionary Hindl-Sabdasdgar,^ and after
him a number of authors, in particular ^ri Jagannath Prasad ^arma.'*
The most essential of all their arguments is firstly that Lallu Ji Lai
did not act independently, but by the direction of the Administration
of Fort William College ; secondly, his weak knowledge of Sanskrit
and insufficient preparedness for such an important business as the
creation of a literary Hindi.
The Hindu scholars point out that besides Lallu Ji Lai and Sadal
Misra the creators of literary Hindi are to be considered Munshi
Sadasukh Lai and Syed Insha Allah Khan, whose actmty in that line
is all the more important as they both acted on their own behalf and
initiative, and not upon the directions of the administration of Fort
William College. According to their assertions this fact is principally
of great importance, as it is a proof that Indian society realized the
‘ Compare also : Syam Sundar Das ; Hindi gadya ke adi dcarya. Xa‘ pra"' Pa°.
navin samskaran, bhag 6, ahk 1, 1982, p. 19.
2 Bhasa-rijhan, pp. 348-9.
® Hindi bhdsd kd vikds, p. 41.
* Hindi kl gadya saill kd vikas, pp. 197-8.
384
A. BAKANNIKOV —
necessity of having a prose literary language ; attempts were made,
independently of the English, to satisfy this need.
Munshi Sadasukh Lai was a native of Delhi. He was born in Sam.
1803. For a long period he was an employee of the East India Company.
He knew the Urdu and Persian languages perfectly and wrote several
works in those languages. He made a translation in prose of the
Bhagavat and gave it the name of Sukhsagar. Besides that, Sadasukh
wrote 1 several articles of which one was even printed.
Ramcandra Sukla emphasizes that Sadasukh wrote in prose
language not according to the directions of some or other Enghsh
official and not according to some kind of given standard, but according
to his own initiative. As follows from the above, Sadasukh wrote in
the spoken language of educated Hindus.
One may doubt the proximity of Sadasukh's language to the spoken
language, anyway as far as his lexicology is concerned, because of his
using Sanskrit tatsamas to a great extent. It is certain, however,
that he followed the same way as most of the workers of the modern
Hindi literature, who also use Sanskrit tatsamas to a large extent.
As Sadasukh began to write a little before the other workers of
the beginning period of the Hindi prose literature, his activity in this
respect was considered to be of great import.^
Syed Insha Alla Khan (died in 1817) was the most eminent poet
of the Urdu hterature. He was a poet at the court of Delhi and after-
wards of Lucknow and finally he was in Murshidabad at the court of
the Nawab of Bengal. Wishing to prove to his friends it was possible
to write in a pure spoken language, Insha AUa Khan wrote a story
called “ Rani Ketki ki Kahani ”, in which he used only the vocabulary
and terms of style of the spoken language of the educated circles of
his time, avoiding both vulgarity and pretentious expressions typical
of literary Urdu of his epoch. Insha Alla Khan’s independence of
topic, the simphcity, refinement, picturesqueness and vividness of
style, imbued with expressions of everyday life, force the historians
of hterature to acknowledge Insha Alla Khan’s pre-eminence in the
way of masterly use of the new prose style of literary Hindi, although
historians generally point out the influence of Urdu in his lexicology
and syntax.
The third creator of the modern literary Hindi the Hindu scholars
consider to be Pandit Sadal Mifra. As well as Lallu Ji Lai, he served
‘ Syam Sundar Da.s : Adhunik Hindi gadya ke adi dcdrya, p. 18.
“ Ramcandra Sukla : Hindi sahilya kd vikds, pp. 210-211.
MODERN LITERARY HINDI
385
in the East India Company in the capacity of teacher at Fort Wilham
College. In the year 1803 he translated into Hindi the tale
“ Nasiketopakhyan Although, according to the opinion of Hindu
scholars, the language of Pandit Sadal IVIisra is much better than that
of Lallu Ji Lai, it nevertheless has some deficiencies, of which the most
important are : The using of the Eastern Hindi forms such as pkulanh,
cahudis, suni, etc. ; placing auxiharies before participles of the
verb, for instance, uttam gati ko hai pahucte, etc. ; the dropping
out of the copula, for example, kanyd sab gati ; instability of
orthography, such as kad hi and kadhi and several other defects.
Notwithstanding these trifling deficiencies, the language of Pandit
Sadal Misra is considered to be better than the language of Lallu
Jl Lai.
Especially Lallu Jl Lai and his language are subjected to harsh
criticism.
Jagannath Prasad Sarma ^ points out that the work of LaUu Ji
Lai is much inferior to the mentioned works of Sadasukh and Insha
Alla Khan, because Lallu Ji Lai acted not upon his own initiative but
according to the directions of others, being in the capacity of teacher
at Fort Wilham College. Besides all that, his “ Prem Sugar ” written
in 1803 at the same time as “ Nasiketopakhyan ” of Sadal Misra and
“ Rani Ketkl ki Kahani of Insha Alla Khan is not considered to be an
original production. The language of Lallu Ji Lai, according to the
opinion of the mentioned author, is subjected to many deficiencies,
of which the chief are : there is no steadiness in grammar, Sanskrit
tatsamas are used to a great extent, but their spelling differs from
the standards of orthography. The orthography of tadbhava is also
not strictly kept to. Lallu Ji Lai has quite excluded Arabic and Persian
elements from his vocabulary, whereas these elements have been
without doubt substantially introduced into the Hindi and, on the
contrary, he often uses Braj and introduces into Prem Sagar
a large quantity of verses in that dialect. In other parts of his works
besides poetry in the Braj language he sometimes makes use of
rhymes.
Based upon the above deficiencies, Lallu Ji Lai s language, which,
according to the opinion of the Hindu scholars is not void of some
positive traits, cannot be considered as an example of hterar\ Hindi.
For that reason Lallu Ji Lai cannot be considered to be the founder of
that language.
' Hindi H gadya saili kd vikds, p. 197.
386
A. BARANNIKOV —
Of the four authors who wrote at the beginning of the nineteenth
century in prose Hindi, according to the opinion of Ramcandra
Sukla,^ the best should be considered to be the language of Sadasukh
Lai, and for this reason he must be considered to be the creator of the
modern literary language. Sri Jagannath Prasad Sarma is of the same
opinion in this respect as Ramacandra Sukla.®
From the point of view of purity and correctness of the literary
language Syam Sundar Das ranks the three authors as follows ; “ The
first place is to be occupied by Insha Ulla Khan, the second by Sadal
Misra, and the third by Lallu Jl Lai.” ^
In this way Lallu Jl Lai, who in the European literature is con-
sidered to be the inventor of modern prose Hindi, from the point of
view of Hindu scholars occupies the last place, or is even completely
excluded from the ranks of parents of the Hindi prose.
In Order to estimate the cause of such varying conclusions both on
the part of European authors and Hindus, it is necessary to dwell upon
the arguments used by Hindu scholars and their criterion of a model
literary language.
Hindu scholars consider the deficiency of Lallu Jl Lai to be in the
following ; his grammar is not stable and has no standard ; for
instance, he uses several variations for the same form ; in order to
express the Conjunctive Participle he uses such forms as : Jcari,
karke, hulay, bulaykari, bulaykar, bulaykarike. Further, in Lallu Ji
Lai's works we come across Braj dialect, such as : bhal, sm, nirakh,
lljai, and others.
Diversions from the standard modern language are to be found in
the works of all four authors. In Sadasukh Lai we come across such
forms as ; dvtd, jdvtd, etc., ko instead of kol, etc. ; Insha Alla Khan
uses such forms as dtiyd , jdtiyS, etc. ; Sadal Misra employs forms
already noted like phulanh, etc. Thus this deficiency is observed
in the works of all the mentioned authors. It is doubtful
whether one should reckon these deviations as a deficiency, for
most probably the spoken language of the latter period of the
eighteenth century possessed a greater number of forms than
the modern literary language, and different authors introduced
variations of these forms into their works. Only the subsequent
development of the literary language actuates a certain standardization.
^ Hindi mkitya kd vikn.% p. 214.
^ Hindi kl gadya .saili kd vikdSy p. 195.
® Adhunik Hindi gadya ke ddi dcdrya^ p. 33.
MODERN LITERARY HINDI
387
Secondly, even at the present time, the literary Hindi is not
standardized to a great extent among different authors, not mentioning
noticeable discrepancies both from a lexical and grammatical side.
As an example we may take the forms of the Conjunctive Participle,
for which any grammar provides several variations, for example bol,
bolkar, bolke, bolkarke, or the Conjunctive form jde, jdye, jdy, jdve,
or the form of the Past Participle such as diye, die, etc.
There is no doubt, of course, that some of these variations
will gradually drop off with the further development of the literary
Hindi.
The fact of Lallu Ji Lai employing verse in the Braj dialect can by
no means be considered as a deficiency of his language. The adoption
of verse in the Braj should be looked upon as a peculiar method of
composition practised by Lallu Ji Lai in his “ Prem Sagar ”. Epic
narration is carried out by him in prose language, moments of high
lyrical tension are reproduced in a lyrical language, which the Braj
dialect was ages ago considered to be. Neither of these dialects are
mixed up with each other. Only single forms of Braj are sometimes
used in prose, which can evidently be explained by the fact of their
being in affinity to the spoken language of the end of the eighteenth
century.
Hindu scholars consider Lallu Ji Lai's language to be greatly
deficient by his using Sanskrit tatsamas differing from the orthography
generally used.
Evidently Lallu Ji Lai, attempting to write in a language the nearest
possible to the spoken, without doubt wrote Sanskrit tatsamas in
such a way as they were pronounced at that time. It is a defect of
orthography, but not a defect of the language.
Orthographical questions are very complicated ones, and Lallu
Ji Lai, creating a new literary form, solved the difficulties as
he considered to be more correct. Secondly the orthography
of the literary Hindi is its weakest point, and is to be further
improved.
The Indian Press has recently discussed various projects of
reforming the devanagari. In case one of these projects should be
accepted, naturally the orthography of many modern authors will
become obsolete, but this does not mean that their language will
become obsolete as well.
Because of the digressions from the Sanskrit tatsama orthography
many authors state that Lallu Ji Lai had quite a low knowledge of the
388
A. BABANNIKOV —
Sanskrit or even did not know it at all.^ I doubt whether such a
conclusion will be correct after a minute survey of the language of
“ Prem Sugar Such conclusions should not be drawn from this fact,
all the more so as Lallii Ji Lai was not prepared for such a role as the
creator of a new hterary language.^ Syam Sundar Das who also con-
siders Lallu Jl Lai hardly prepared enough for his role, still brings
forward the opinion of other scholars ; “ some say if he lived at the
present times he would never have attained such fame. But this may
be said about Newton and other world famous scientists.” ®
Besides, the role in history of this one or another promoter is not
determined by the degree of his erudition. Most probably at the end of
the eighteenth and at the beginning of the nineteenth century there
were many scholars who knew the Sanskrit and most likely the Hindi
much better than Lallu Jl Lai, but, nevertheless, they did not
participate in this great undertaking as Lallu Jl Lai did.
Hindu scholars, basing their views upon abstract facts of purity
and correctness of language, find it impossible to consider Lallu Ji
Lai the founder of the modern literary Hindi, and they consider that
such a role should be attributed to Munshi Sadasukh and Syed Insha
Alla Khan and to a certain extent to Sadal Misra.
One may come to such a conclusion only in that case if one is to
forget historical facts announced by these scholars. It is well known
that the story of Sadal Misra, “ Nasiketopakhyan ”, was soon forgotten
and was not republished, ^yam Sundar Das himself says * the collection
of Sadasukh’s articles was not even published, and the story of Syed
Insha Allah Khan ® was published for the first time by Eaja Siv
Prasad. Whereas “ Prem Sagar ” by Lallu Ji Lai was published many
times and its popularity grew stronger and stronger. It is quite com-
prehensible that the formation of a literary language should be affected
only by such works as are propagated and widely read and not the ones
which lie in manuscripts or which are forgotten inunediately after
their appearance.
For this reason from all the mentioned works of the four authors
considered by the Hindu scholars to be the founders of modern literary
Hindi, the “ Prem Sagar ” of Lallu Jl Lai is the most noted work to
play an honourable part ; owing to it the idea of a prose literary
* Raracandra Sukla : Hindi mhilya ka vikan, pp. 212-13.
“ Syam Sundar Das : Adhunik Hindi gadya ke ddi dcarya.
’ Op. cit., p. 30.
* Op. cit., p. 18.
^ Op. cit., p. 32.
MODERN LITERARY HINDI
389
language became popular, many authors of the following stages of
the development of literary Hindi studied prose language by it.
Hindu scholars attempt to di mini sh the significance of Lallu Jl
Lai and Sadal iVIisra by pointing out their serving at Fort William
College and their acting upon the initiative and directions of the
administration of the College, whereas Pandit SadasuMi Lai and Insha
Allah Khan were never employed at that College and therefore acted
independently and upon their own initiative. But, first of all, Pandit
Sadasukh also served in the East India Company, although not at the
College ; secondly, the fact that both the mentioned authors started
to work upon a prose language only when Lallu Ji Lai and Sadal
Misra were solvdng the same problem, and this undoubtedly is a sign
they began their work also under the influence of Europeans.
The influence of Europeans upon the development of a prose
language is not only to be seen in that they brought to India an idea
new to this country of a literary language resembling that of a spoken
language, but also as Sri Jagannath Prasad Sarma ^ justly remarks,
with the advent of the English and under their influence great changes
have taken place in the economic, social, and religious life of India.
A new bourgeois class is formed which is in need of a literary language
close to that of the spoken, and, with the assistance of the Enghsh,
this class creates and spreads it by technical means (such as the press,
etc.) adopted from Europe.
Out of three literary forms set up on the basis of Kharl boll, viz.
High Hindi, Urdu, and Hindustani, this latter the Hindu scholars
consider to be purely artifi cial, originated by the English for pohtical
reasons These scholars imagine the affair to have been enacted
in the following wav : the English selected out of the Urdu and Hindi
words common to both languages, kept the Hindi grammar and in
this way invented a new language. Such a mode of explaining the
formation of the idiom which the Hindus themselves call Hindustani
is quite mechanical and contradicts their own words, as the Hindu
scholars declare that the Hindustani is just the form of language
jise sab log bolcdl me kdm me late hat ^ “ wUich is used by all in
conversation ”.
From the point of view of proximity to the living spoken speech
the Hindustam- is the most perfect form of a literary language which can
1 Hindi ki gadya aaili ka vikds. pp. 189-190.
§vam Sundar Das : Bha<a-rijndn, pp. 342-5.
3 gyam Sundar Das : Bhasa-fijndn, p. 342, 1.
390
MODERN LITERARY HINDI
unite the Hindus and Muslims. This is recognized by several Hindu
scholars, for example by Baburam Sabsena.^
Most of the Indian scholars have another point of view upon the
matter and find that the hterary language must differ from the spoken
language, even of the educated class ^ ; in accordance with this, these
scholars see the only way of forming a hterary Hindi on Hie basis of
a spoken language by satiating it with as many pure Sanskrit tatsamas *
as possible.
The satiation of hterary Hindi with Sanskrit tatsamas is not only
done out of “ purely theoretical considerations, but with the purpose
of rendering the Hindi comprehensible in other provinces, as in the
hterary languages of these provinces a great number of Sanskrit
tatsamas is also found.
It is omitted in this reasoning that the spoken Hindi or rather the
Hindustani did not need Sanskrit tatsamas in order to be widely
spread.
Although a sound judgment is raised in objection to excess of
Sanskritizing the Hindi, for example by Ham Das Gaur * and others, also
several modern authors of literary Hindi Sanskritize their language
to such an extent that tadbhava elements occupy quite an insignificant
part in their vocabulary. For instance, in Viyogi Hari ® we read :
jab tml . . . srotasvati-sarit-tai-tani-sakhd-vikarit-kalJcanthi-kokil-
kuhuk-dhvani sxintd hu, 'prahhdt-ausakati-jhalkit-harit-trndcchadit-
'prakrti-'pariskti-bah u-vanasfati-sngandhit-sukhad-bhumi far lettd hU.
From the above specimen w’e may draw the conclusion that a
digression from the principle of connecting together spoken and hterary
language leads, practically speaking, to the restoration of Sanskrit.
^ Bhdratvars kl ddhunik Ar>ja bJmsde^ Pra^ Pa^y navin samskaran, bhag
11, ank 2. pp. 121-162.
2 Syam Sundar Das : Rhdsd-vijudny p, 353.
3 Mahamahopadhyaya Sri Giridhar i^arma Caturvedi ; Varttamdn Hindi me
Sarhskrt sabPd ka grahan, Xar Pra° Pa^, navin saraskaran, bhag 10, ank 1-2, 1086,
pp. 19, >-231.
* ^rirdmcaritmdtia!^ I'l bhumikdj Dehli aur Kasi, 1982, p. 8.
3 ^rl Jagannath Prasad iSarma : Hindi ki gadgn mill kd vikdSy p. 334.
Bi ll. S.O.S. Vol. VIII. Parts 2 axd ;5.
Plate III.
Pahlavi Fralmext T.M P.C) (1^ 1).
'! uf.K.
Remarks on the Pahlavi Ligatures ^ and ^
By K. Barr
(PLATE III)
I N the Sitzungsherichte der Kgl. Preussischen Adademie der Wissen-
schajten, Jahrgang, 1904, pp. 1136-7, K. F. Geldner published
a transcription of a fragment of a Pahlavi-frahang found in Turfan
and now preserved in the Museum fiir Volkerkunde in Berlin
(registered as TM 195 (Pi) ). As some graphic peculiarities and
especially the ligatures occurring in this fragment may be of particular
interest for the history of the Pahlavi system of writing, I think it
may be worth while to reproduce a photograph of it, obtained through
the kind assistance of Dr, W, Henning and Dr. Gelpke, Berlin. For
permission to use it for this purpose I am greatly indebted to the
kindness of the Secretary of the Preussische Akademie der Wissen-
schaften, Geheimerat Liiders.
The fragment contains part of a list of verbs. This list gives far
more grammatical forms of each verb than the Pahlavi-frahangs
published by Junker. On the other hand, it does not give the Iranian
equivalents of each verb-form, but only, after the enumeration of the
ideographically written forms, the corresponding Iranian infinitive.
Apart from the ligatures, the discussion of which is the main purpose
of this article, the fragment has the following peculiarities. The
infinitive is always written and the past partic. both without
the perpendicular stroke generally found added to these forms in
Book-Pahlavi.^ The stroke (marked ‘ in my transcriptions) is found
' I take it for granted that this stroke in Book-Pahlavi represents the final -y
of the Sasanian inscriptions and the Pahlavi Psalter, found not only as the ending of
the eas. obi, of nouns, but also, being part of Aramaic forms, in ideograms like
L‘YNY and 'YTY ~ hnut, L'YTY ~ nest, which words in good MiSS. always
I may mention here that the - 1’ of PA T has been
are written
convincingly explained by Andreas as the .\ramaie dual-ending. This perpendicular
stroke is, in my opinion, the only trace of the eas. obi. found in Pahlavi of the books.
It is, of course, only a graphic .survival, and I consider the of the inscriptions and
the Psalter to be so too, because only on thi.s assumption is it possible to account for
the irregularity of the -y being used or omitted.
392
K. BABR —
only in the ending -isn, written -hi' (R 1, 5, 10 ; V 1, 5, 10). At the
end of a line some letters have special forms : y -yy, the ending of
the 2nd sg. pres. ind. (R 12 ; V 7), ^ (R 4, 13 ; V 4). The letter n
sometimes has a shape with a curv^e below, which recalls the ^ of
the inscriptions, the - j of the Psalter. This is the case at ends of lines
R 6, 7 (perhaps also R 1), but also in initial position V 7),
and sometimes in the compounds ^ ’n R 2) and u (v. the
—
forms of ))^^y V 8 seq.). This special form of n might, of course,
as well be considered merely a peculiarity of the individual hand.
The other graphic traits of our fragment, such as the diacritical marks
_ and marking i as respectively y oi d (wrongly used in a few forms
of YBLWN V 1, 2) are met wdth also in the oldest Pahlavi MSS.
known.
In the fragment the following seven verbs are represented : I,
dnitan ; II, daMan ; III, sutan ; IV, hurtan ; V, nltan ; VI, [uzitan] ;
VII, [vinddtan]. As the only form preserved of VII is the imper. sg.,
I omit this verb in the synoptic list I now have arranged in order to
make out the value of the ligatures of endings found. The ligatures,
which will be discussed below, are put in parentheses, restorations
in brackets.
I II
III
IV
V
VI
(3) TffSX.Y
‘ZLwy
[YBIWX]
*DBLWN
rvp.fi: ir.v
— cl
(S)—d
[r]BLWXd
(3)— (?)
— d
— yy
— yy
YBLXyy (!)
— yy
— yy
— yt
— yt
(1.3) YBLWXyt
— CO
(8) — yt
(4) -
— 't
[-
(4) — [vt)
— ’t
— m
(9) — m
r
( — m
— w
— m
— yc^
— ym
(14) — ym
— ym
(9) — ym
(o) — 'm
— ’/«
YBh\ )r.V’»q
— 'm
— ’«(/
(10) — ’ml
r YBL irv’«ir/
{.■>) —
— ^nd
(R 1) HyTYWXm' — sn[']
— in’
(V 1) — ,sft'
— .y/fc'
(10) — sn'
— [yilyi (6) — yhi/t
— yHyt)
— yhyl
— yUyt)
— yhyl
— ([ft ?] — yh-it
(11) — yh.it
— yh{st)
(6) — yhst
—
(2) — yht — tn
— In
(2) - tn
— ((ft)
(11) — tn
— t (7)— I
— t
— t
— t
— t
'nytn d'itn
(12) iui[n]
bwrtn
(7) nytn
[ 1
A glance at this synopsis shows that, with one exception only,
each verb is represented in the same grammatical forms, only with
a few’ discrepancies as to the order in which the forms occur in the list.
II, III, and IV are in complete accord, and IV had very Ukely, at
REMARKS ON PAHLAVI LIGATURES
393
least originally, the same number of forms agreeing with the others
wherever a trace of a final is left, thus making the restoration of the
form in question Cjuite certain. In Y the 3rd pi. conj . i.s missing, and
the indie, and the conj. 3rd sg. are interchanged. In I, too, an inversion
of forms has taken place, that of the infinitive and the 3rd sg. pret.
pass., if my restoration of the text is correct. As for
R 1. exactly fills the gap, and at the end of the same line I think
I discern a trace of the final n (cf. at the ends of lines 6, 7) making
up the infinitive. I am, of course, not quite sure of the last point,
as I have not had the opportunity of examining the original manuscript.
R 10 and V 5, the ending -yt of the 3rd .sg. pres. ind.. elsewhere
written with the common has a shape that cannot well be separated
from the final ^ of the MSS. The same form of the final with the
great loop and the curved down-stroke, occurs in the compound -si
in the 3rd sg. pret. pass. YBLWXyhst V 1, elsewhere in the fragment
written in the common way At the end of Y 3 we find a ligature
which I hardly think can be read otherwise than 't, the ending of the
3rd sg. conj., elsewhere though we have to assume the above-
mentioned inversion of the 3rd sg. conj. and the 3rd sg. indie. The
second part of this ligature has exactly the same shape as that of the
ligature of the JISS. It is to be noted that these ligatures are all
used at the ends of lines, but whether this fact is more than accidental
we have no means of deciding, owing to the very limited extent of the
fragment. However that may be, the shape of the especially in
R 10, Y 1, 5, might easily be explained from the use of peculiar
final letters in different kinds of book-scripts of Semitic origin.
In Y 6 we find another ligature, which according to the parallel
forms of the other verbs surely represents hi. The resemblance of this
ligature and the ^ of the MSS. is. I think, so striking that the two
signs must be connected. And I think the shape of the ligature in
our fragment gives us the clue to understand at least one very frequent
use of The ligature in V 6 contains the same form of found
in the ligature 7 Y 3. and the form of n mentioned above, p. 392.
As for the cross-stroke found in this ligature and in the ligatures for
-yt and -7, 1 am not sure how it should be explained, but I feel inclined
to regard it as an ornamental element, perhaps serving the purpose of
binding the elements of the compound together. It is found not only
VOL. VIIT. PARTS '2 AND 3.
394
K. BARR —
in the ideogram ^ andar < BYN, but also in the Av. ligature
perhaps introduced there from Pahlavid
Another element, the interpretation of which is somewhat doubtful,
is the small oblique stroke added to the ligature -yt in R 10, but not
in V 5, and perhaps to the ligature 't V 3. We may, however, compare
the perpendicular stroke very frequently, but not regularly, found
with the endings -yt and -'t of the 3rd sg. pres. ind. and conj. in our
MSS., cf. the -y of the personal endings of the Psalter (the inscriptions
only have -my of the 1st sg.). The absence of the stroke in the ending
-yhst V 1 is in accordance with its omission in the past participles
in the fragment.
Finally we have to discuss the bgature with which the first form
in V 3 is supplied. From the parallelism with the other series I conclude
that the form in question is the 3rd pi. indie., being in the other verbs
always characterized in the common way by adding 5 to the ideogram.
This ligature is perhaps to be explained as representing -ynd, i y
being combined with a ligatiue of d and the aforesaid form of n.
If this explanation is correct we must assume that the ideogram in
this case has been supplied with that form of the ending, which is
normal only in verbs not ideographically written. I have met such
spellings in a few cases in the old MS. of the Vendidad K 1. The
instances found in this MS. are, however, not conclusive because
may be only a clerical error for
Provided that the explanations given prove to be correct, it can
be stated that our fragment uses the ligatures, which I connect with
and of the MSS. respectively, in a way which is very common
in the Book-Pahlavi. The cases in which forms in alternate with
^ and forms in ^ with are in the MSS. so numerous that I cannot
agree with Xyberg. who regards every other use of the endings than
that which he has tried to establish in his Hilfsbiich as only being due
to errors of scribes who did not understand the real value of the
1 It alway.s ought to l)p kept in mind in discussing the origin of letters of the Av.
alphabet that Pahlavi and Avestan writing was executed by the same scribe.s. Thus
we .shall have to take into aceount the fact that the form of the Av. script, such as
we know it, might be due not only to an '■ I'r- "Avestan Aramaic .script, but also,
to a certain extent, to a .secondary influence from scribes versed in writing Tahlavi,
the si ientitie language of Zoroa.“trian theology (of. e.g. the formal eongruitv of Pahl.
M and Av. M a which cannot be anything but secondary).
REMAEKS ON PAHLAVI LIGATURES
395
ligatures (Hilfsbiich, i, Einleitung, p. 18). As there are some texts,
as, for instance, the Menok i Xrat of K 43, which do not use these
ligatures at all, whereas they occur very frequently in others, the
question certainly needs closer examination, which must take into
account also the problems of syntax and chronology of the respective
texts and MSS. I must here confine myself to a few references.
for -yt, the ending of the 3rd sg. pres, ind., A(^iyatkar i)
Z(areran), §48, end: J
^ j where a few lines above
in the text the exactly parallel passage has instead of
As for similar cases, cf. §§ 51, 74, 76, 83, 88, 92, 95, 103, 105,
106, 110, 111 of the same text. for -yt in the 2nd pi. imper. is
found, e.g. A.Z., § 41 : ^ maiestatis on
account of hnah bayan being the subject). As for the interpretation
of §§ 6 and 7, I am not so sure, though I consider a reading
hilet not altogether excluded. A certain instance of hilet written with
is found in § 96. Further examples of the 2nd plur. imper. are
foimd in Kn. (Sanjana), idii, 12, Artaxfir . . . f ram fit ku ataxs kiinet
(j|J5°) “ A. ordered : Kindle a fire ! " ; viii, 7. framilt ku ka ac diz e
kirm dilt tenet, martdnaklh . . . kunet (^°) b bun i diz dyet (^°)
“ when you observe the smoke from the citadel of the worm, be
courageous and come up against the citadel ! ’’
^ occurs in the MSS. frequently for the ending of the past
partic. (and, more rarely, for the of the infinitive), the > in
the MSS. representing the y-stroke and not the n as in our frahang.
This mode of noting the past partic. is extremely common in the MSS.
of the Avesta-translation and especially in the Madiyan i hazar
dadastan, and is not rare in other Pahlavi-writings. Cf. A.Z., § 2,
patgrift ; Kn. viii, 2, stat, 3. patgrift kart ; ix, 11, stat
with v.l. ^°, x, 15, nit. v.l.
I do not see any difficulty in explaining ^ and ^ from cursive
y + t and t n. I suppose the ligature to have been conceived
rather as one single character than as a compound letter, and this
396
K. BARR —
to be the reason why the shape of the t has lost its proper and
distinct shape in . The same may happen in the case of
being in MSS. rather frequently confused with ^ (as is also ^ with
y). As for the curve of the down-stroke of the I think it has to
be explained from a merely aesthetic point of view as due to influence
from the left-hand curve of It is not likely that this shape has
anjdhing to do with the Sasanian form ^ (vv infra, p. 399. footnote 2).
The fact that the ligatures ^ and ^ are used in the above
way in a frahang makes it, I believe, almost certain that we have to
do with a generally adopted system of orthography. A frahang
arranged like ours has most likely not been composed to facilitate
reading of Pahlavi but to teach how to write it. Otherwise it would
certainly not have given so many verbal forms without stating how
they are to be pronounced. It seems to me to be an orthographic
handbook teaching how the most important forms of the ideo-
graphically written words are correctly spelled. So I cannot help
thinking that our fragment has some authority and it would be
interesting to know how old it is. Geldner in his above-mentioned
paper from 1904 only remarks : '• Wohl alter als dreihundert Jahre.”
Judging from the character in which it is written, we might suppose it
to be even older. On the other hand, there is no reason to consider it
older than our oldest MSS.
The ligatures ^ and ^ denote in Pahlavi. as is well known,
besides the forms already treated other forms of the verb, not
rarely denotes the 3rd pi. pres. ind. (in this case the MSS. often have
-5^ or ^), and also ^ occurs in this form. It is difficult to say
whether the ligatures are used here wrongly or whether we have in
these cases to suppose that originally different ligatures have coalesced
into those now exclusively found. As for ^ ~ i/n<l we might compare
the ligature of the frahang V 3. This ligature might very well in time
have become .so like the of V 6 that they were confused. This is,
indeed, very uncertain and cannot strictly be proved. The extreme
difficulty of investigations of this order is greatly increased by the
corrupt state in which mo.st of our Pahlavi texts have come down to us.
Thus ^ and are frequently interchanged. We find ^ instead
REMAEKS ON PAHLAVI LIGATURES
397
of in cases where exactly parallel forms in the same clause show
the past participle (as for the change of and cf. the similar
one of ^ and ^ often found in MSS. as Tavadia in his review
of Nyberg’s Hilfsbitch, ZII., 9. p. 276. has justly pointed out), or forms
in in parallelism to forms in or to the naked ideogram
denoting the 2nd sg. imper. The same variations occur in the case of
variue lectiones. We cannot but recognize that the scribes have been
utterly in doubt as to the real value of the ligatures.
Professor H. S. Nyberg has, in his most valuable HilfsbucJi des
Pehlevi (i, Einleitung, pp. 13 sqq.), made an ingenious attempt to explain
the origin and use of the ligatures under discussion. As the adoption
of Nyberg's explanation would involve far-reaching consequences
in our conception of Pahlavi morphology and synta.x, I consider this
sufficient to justify an e.xamination of his views.
According to Nyberg. the source of has to be sought in the
Arsacid letters <h(, that of ^ in the Sasanian ^ O, both meaning
yh. There is, however, as already stated by Tavadia (1.1.. p. 276).
one serious objection to be made to this suggestion : the letter h ,1
is only found in Aramaic elements. Furthermore, we are hardly
justified in assuming any influence of the Arsacid upon the Sasanian
script, and Nyberg has not trieil to give any evidence for this
assumption of his. Though I cannot, for the reason mentioned, adopt
Nyberg's solution of the problem as to the origin of our ligatures, I
think it is not useless to discuss the grammatical side of Nyberg's
theory in the light of the middle-Iranian texts which have been made
available since Nyberg's book appeared, viz. the Turfan texts, ed.
by Andreas-Henning.i and the Pahlavi-Psalter.^
Nyberg sees in the forms in ^ ° the 3rd sg. pres, optative
corresponding to optatives like HW'Yndy of the inscriptions and
'hyndyyh of the Turfan texts. Hence according to Nyberg.
is to be read Jcu»e»deh, the ideogram denoting not only the
pres.
^ Mittelirani-^che Ma)ikhak'(i aus ChinefUSck-Turlf'fifjiK i-ii (south-west dialect),
lii (north-west dialect), von F. C. Andreas (|). Aus deni Xachlass herausgcgeben von
Walter Henning, SPA M'. Phd.-lliA. Kl. 1932, 1933, 1934, quoted as Andreas-Henmng,
i, ii, iii.
* Bruchduckt einer Pehlevi- V bersetzung der Psalmen von F. C. Andreas (f). Aus
dem Nachlass herausgegeben von Kaj Barr. Phil.-Hist. Kl. 1933, quoted as
Andreas-Barr.
398
K. BARR
stem, but also tbe pres, partic. in -nd-. The form is explained as
having arisen by way of analogy with kart -eh. This should be kart he.
or kart-e, the ending normally being written y. This conjecture
must be rejected for several reasons, the first being this, that the use
at the pres, partic. in -nd in Book-Pahl. and in the Turfan south-western
texts is limited in the same way as is the case in NP.. i.e. to that of
a noun, and it is highly improbable that the naked ideogram should
be used to express a form which, though being from a historical point
of view a derivation of the present-stem, does not strictly belong to
the conjugation of the verb.* In the second place, the form HWYndy.
the only one occurring, is found only in the Arsacid parts of the
inscriptions. In the Sasanian parts the corresponding form is written
HWH.^ Thus we must conclude already from the inscriptions that
the form belongs to the north-western dialect. That this is really the
1 It may have been otherwise in the old north-western dialects and Professor
tsyberg's e.xplanation of the north-western optative 'hyndyyh, etc., as a periphrastic
formation from the pre.s. part, may well prove to be correct ; thi.s formation perhaps
survives in the periphrastic pres, indie, in -«(«)- in Zaza, in -nd-. -«(«)- in Sangisari.
LasgirdI, Samerzidi, Talisi, and other dialects spoken in the vicinity of the Caspian
Sea, cf. Liiigvi.stkredsen i Koebenhavn, Aarsberetning for 1934, p. 15, and
Christensen-Barr-Henning, /miiisrhe Dialektaufzeichnumjen au-s dem Xacklas-f von
F. f. Andretti, i, p. 163, footnote 1 (in press). Professor Xyberg told me bj’ letter that
he himself had explained the nd- formations of the modern ■’ Caspian ” dialects in
the saroo way as I did.
The corre-spondence of the north-west and .south-west forms is quite clear from
the Hajj. -inscription. IntheArs redaction (9— 10) we have the past partic. ~ H\V Yndy
(as for the reading, cf. Turfan hyndyyh) in the protasis and the apodosis of an irreal
period : 'YK 'k ayty BFYt HWYndy, . . . 'L-y YHWT HWYndy ' (he said), that if
a construction h.td been erected. ... it would have been visible. " The corresponding
period in the .Sas. redaction (9-10) runs : YK HT cyfky cyty HWH, DYX . . . pty'k
I ^11 11 A H\\ H. Herzfeld reads ast for HW H as W'e. indeed, may do in places where
the Ars. text has H\\ 'it. I should rather sugge.st that HWH here is to be read he
the opt. sg. ,3rd. cf. y ^yij (also or found in the sg.) common in
the irrealis of the past (v. Bartholomae, Zur Kunde d. Mittedir. ilundarten, i, pp. 47-51.
e.sp. p. 50^). Cf. Pahl. Ps. 123, 2. HTmn L' MH MRWHY YK'YilWM HWHd
. . . (3) 'DYSau zywndky ' up rly H W H m " If the Lord had not been with us . . . the.v
would have devoured us alive . P.s. HW Hd mu.st be compared with (not
with the conj. 3rd sg. Turfan h'd {hriS) also used in irreal clauses, as I did in
111 } glossaiy, Andreas-Barr, p. 1.30i7. It is not likely that the spirant 8 <c / in this form
should have been written phonetically with i. i here as in other cases in the Ps. is
for y as in the book-form). The orthography of the inscriptions being rather sparing
as to the use of phonetic complements denoting verbal endings, I think a form like
HWH without any phonetic complement may admit of ihore than one reading. The
iiorth-west texts published by Andreas-Henniiig offer .some instances of this optative
( hyndyyh, icrdyndyh, qryndyh, bwyndyh). We find the same formation of the irreal
EEMAEKS ON PAHLAVI LIGATUEES
399
case is confirmed now from the texts pubfished by Andreas-Henning.
In the south-western dialect from Turfan the 3rd sg. pres. opt. ends
in -e, written -yy, -yyh, -yh {beh, he, sdye, bareh). in the same way as
is the ending of abstract nouns (cf. Henning. ZII.. 9, pp. 234 and
235 Bern.). The h sometimes found might be due to analogy with the
ending of the 2nd sg. pres. ind. At least it can hardly be phonetically
explained if we start from old Iranian optatives in -ait.^ These
optatives are represented in Book-Pahlavi by^^y*) he and
este (Denkart and Av. translation), and I consider it not a priori
impossible that the forms in in some cases may be interpreted
as opt., hke those found in the Turfan texts. If that be the case I
should propose to read the ligature yh, h being explained in the same
way as Turfan -h. Cf. the different spellings of the ending of the 2nd sg.
pres. ind. For the shape of h may be compared the
older form of the h occurring in the Psalter with a down-stroke
on the left hand. Unfortunately there are no examples of a in
final position. As it, however, is very uncertain to what extent the
special features of Pahlavi letters in the Psalter MS. may have existed
in a period prior to the development of the cursive of the Zoroastrian
books, I should not emphasize this weak possibihty.^ It could as
period as in the Hajj. -inscription in a 49-50 (cf. also 6 117. 128). In other cases it is
used to express a wish : ’Jryd bicyndyh “ benedictus sit ” {m 50, 53) = south-west
'fryd byh, or an obligation : (b 53-57) where wrdyndyh and qryndyh are in parallelism to
n.id i'r “ you shall sit down ! ”.
^ As for beh, he we might perhaps think of a connection with O.P. biyd and Younger
A%-. forms like hyat with mood-sign -iyd of the unthematic type.
- For explaining ligatures found in the cursive script the most safe and methodic
procedure is, in my opinion, to start from the cursive forms of the elements forming
the ligatures and not from the forms of letters such as they appear on the stone
monuments or in the Psalter MS. vtTitten in archaic script to serve as a liturgic book of
the church. It is, e.g., quite clear that andar cannot have originated directly
from forms like of the inscriptions or of the Psalter, but only from
: ligature
BY
X. Ligatures, of course, arise in cursive wTiting. and
the ligatures found on monuments, e.g. on the Derbend inscriptions published by
Professor Xyberg, or in the Psalter, are due to the influence from the cursive script
which in Persia as everv-where has developed independently and apart from the
monumental script. I therefore consider it a rather bold undertaking to try to make
out the phonetic value of an obscure ligature in tracing it back to the monumental
script, especially if the forms in question are not found in the inscriptions or in the
Psalter. The ligatures of the cursive are generally as ambiguous as are the single
letters of the Pahlavi alphabet. Thus we always have to consider the possibility of
more than one solution of a Pahlavi-ligature, as we, e.g., in the case of 03 °, may have
to assume a threefold origin : BY + -V, Y + T, Y P H {!).
400
K. BAER
well be suggested that the final
'.L
is due to influence from the
originally different ligature meaning -yt. In fact, we rather often find
the ligatures and written with a ^ almost identical
in shape with the Av. final ^ (fairly often in the MS. K 20).
The possibility of the occurrence of such optatives in being
admitted, it must be stated that it it very difficult to ascertain the
exact extent of their use. In the Turfan texts written in some south-
western dialect the optative is used to express a wish (so always
hyh ; hy in M. 219 E 7 Andrea s-Henning ii) and as a potential mood
(M. 49, ii, V 10-15 : M. 9, i, E 12-16 I.I. ; cf. Henning, ZIL, 9, p. 2.36).
In Book-Pahlavi it might be found as a potential in some kinds of relative
and conditional clau.ses, but on account of the incertitude of our
manuscript tradition we can never be sure. The whole problem must
be taken up in connection with a treatment of the use of moods in
general, and I hope in a not too distant future to be able to publish
the results of my studies in this field, and shall therefore here restrict
myself to a few remarks.
The optative mood can, apart from the form he, not have been a
real living form of the verb in the language at the time when the
redactions of the oldest MSS. took place. That is perhaps the reason
why we never find a verb not ideographically written in the optative
mood expressed with That only the relatively few ideo-
graphically written verbs enumerated in the fist, Hilfobuch, i,
Einleitung, p. 14, to which sutan, frequent with
in the Av. -translation, may be added, should have preserved the
optative, if it really was a living mood, would be very difficult to
believe. I see in the few cases in which ^ ° is not likely to represent
^ (or some other form in cases of wrong use, e.g. = -ynd
Yd. 4 i5 ^ 5^^, = I : ~ Av. conj. frabardl Vd.
hit Vd. ISg, etc.) only graphic survivals
taken ov'er by the later redactors or compilers from their sources.
The ligature ^ ° (and the same assumption may perhaps hold good
in the case of as a mark of the imperative 2nd sg. ; cf. infra) could
be preserved when combined with an ideogram because the copjdsts
took those word-pictures as totalities representing to their minds the
KEMARKS ON PAHLAVI LIGATURES
401
pronunciation of their own time, while the phonetically written verb
forms were brought into accordance with the spoken language in which
the role of the optative mood had been taken over by the conjunctive
and, to some extent, by the indicative. Some optative forms seem
to have disappeared early in Zoroastrian Pahlavi. Thus we, e.g.,
never find, as far as I know, a * which would correspond with
Turf, hyh, but only bavdt or bat in the same function as that of Turfan
byh, and in different MSS. we find hat used in the same way as he,
a development which may have started earlier in the dialect, which is
at the base of the Sasanian Koine, than in the south-western dialects
of the Turfan texts ; in the Pahlavi translation of the Avesta we have
fairly often the 3rd pers. indie, with the particle ’y{w) (Turfan hyh) ;
the same is the case in the Psalter while the Manichsean Turfan texts
in a few cases have preserved the optative (cf. Andreas-Barr, 120a
s.v. VX and Henning. ZIL. 9, 248).
In order to get a clue to the interpretation of the ligature
Nyberg starts from the correct observation that it is very frequently
found added to ideograms, where there can be no doubt that the
meaning is that of a 2nd pers. imperative. Nyberg reads the ligature
eh, which, according to him, is to be derived from old Iranian medial
imperatives of -aya-stenis, showing -ayahica in the sg., -ayabvam in
the pi. These two endings coalesce into Pahl. -eh. thus being the ending
both of the 2nd sg. and pi. imper. As to the pi., Nyberg finds a support
to his view- in the 2nd pi. imper. hyrzydw of the Turfan texts. I think
this form had better be left out of consideration. It is only found once
(M. 4a, 14, cf. the remark of Salemann. Man. Stud., ii, 167), and it is,
if at all genuine, a north-western form, hence not being really
conclusive as to the south-western dialect. The te.xts published by
Andreas-Henning have only -yd in the pi. in both dialects. In Book-
Pahlavi we find and, with ideograms, not rarely ^
cases where from the context we expect a 2nd pi. imper. may be
explained from the frequent confusion of and ^°. Thus, in
my opinion, only as a mark of the 2nd sg. imper. needs to be
discussed here.
As stated above, p. 397. 1 cannot adopt Professor Nyberg s reading
and explanation of the ligature, but I must immediately confess that
^ Cf. now Henning, 1935, p. 7, footnote 3.
402
K. BARR —
I myself have not been able to find a satisfactory solution of the riddle.
Any evident analysis of the ligature is only possible if we can prove
it to be an adequate expression of a really established ending of a
middle Iranian 2nd sg. imper. The possibility of the survival of the
imper. middle is in my opinion very weak, as the middle voice no doubt
already at the end of the old Iranian period was gradually disappearing.
At least the existence of a middle-ending -ayahva cannot be proved by
any analysis of the ligature The only possible coimection I can
think of is that of the ending -e < -dija. sometimes found in Book-
Pahla\’i (cf. Tedesco, ZII., 2, 306 tf.). Unfortunately the inscriptions
give no evidence, and the evidence of the Psalter in which we meet
imperatives 2nd sg. supplied with a -y (hivc-y. Jc'm-y. 'mwc-y) is rather
weak, because the Psalter in some cases seems to supply verbal forms
with -y without any phonetic or morphologic ratio, e.g. 1st sg. pres,
in -my, 2nd sg. in -ijdy (i.e. y + y, I cannot adopt the view of
Henning ; "l pseudohistorische Schreibung fiir gesprochenes h ’’
ZII., 9, 236), 3rd. pi. in -yndy ; those forms occur only with verbs
when written phonetically, the orthographic use of -y with verbs
following the same rule as the -y of the cas. obi. of nouns !).' The
ending -e is written y° -yy or -yh, in other cases we find also
-u»°, yt°. There seems to be some confusion, on one side, with the
ending of the 2nd sg. indie., on the other side perhaps with the 2nd
sg. conjunctive -'y known from the south-western Turfan texts and
from the Psalter = 'h of the north-western texts and in both dialects
often used as an imper. (in this way I am inclined to interpret the
forms in -iu° A. Z. § 92, savdy, d^aray, § 93, hovdy and perhaps
Jramdydy though spelled with § 109 ddrdy ; the ending is in all
those places ascertained by the metre as restored by Benveniste,
JA., 1932, 245 ff. Cf. also Andarz I O.snar, § 48, zut be glrdy " catch
it quickly he spozay “ leave it ! "). Though I cannot prove it, I do
not find it altogether impossible that the ligature ^ may be a
^ The forniri in with pron. suflT. arc not clear to me. A change: -c in final, -n in
anteconsonantic positum ii? not very probable in a text written in purely south-w'cstern
dialect, as the JViIter. A reference to the rather eonfused materials collected by
Lentz. ZJL, 4, 270 ft'., does not help to clear up the problem. Professor Arthur
Christensen once, when I discussed the pnjhiem with him, suggested that ’w, 'mn
might be writings of the pron. auff. comparable with modem Pers, used
after
EEIIABKS OX PAHLAVI LIGATURES
403
variant of y perhaps modified through influence from a similar
ligature originated from ^ ) like that discussed above, p. 395.
If this mere suggestion, as I freely own it to be, should prove correct,
we must conclude that the -aya forms in the Pahlavi orthography
have come down to us in a double form, one in^ in use both with
ideographically and phonetically written verbs, and one expressed
through the obscured ^°, only in use in connection with ideograms
like the optative in There remains, however, a difficult question
to be solved. In cases where a clause contains more parallel imperatives
we regularly find the ideographically written forms in paralleled
with phonetically written forms without ending (cf. Nyberg, Texte
zum mazdayasnischen Kalender, p. 48 : yos roc parvarisn i gos-urvak
kune u gdv 5 varz dmdc). This fact might perhaps be explained thus :
the imperative originated with the -a stems, which has prevailed in the
south-western dialect of the Turfan texts, the Fars-dialects, and the
modern Persian Koine, at an early time entered into competition with
and, at last, superseded the imperative in -e < -dya ; hence the -e
was, as being foreign to the spoken language, neglected by the copyists
in phonetically written verbs. The problem certainly needs further
investigation.
This article was almost finished when Dr. Walter Henning, Berlin,
kindly sent me his review of Nyberg's Hilfsbuch {Gott. gel. Anz., 1935,
pp. 1-19), in which he deals with Nyberg’s explanation of the ligatures
and ^ (pp. 6 ff.). I agree with Dr. Henning in most of the particulars
regarding the interpretation of text-passages, but not with his general
views of the problem under discussion. I therefore pubhsh my article
in the main unchanged, having only left out a few examples which,
in my opinion, have been satisfactorily explained by Dr. Henning
(and partly already by Dr. Tavadia in his review, ZII., vii, p. 273 ff.).
Sur quelques dvandvas avestiques
Par E. Benveniste
T AXT cjue le dvandva n'est pas devenu, comme c’est le cas dans
I'histoire de Findien. un procMe de style, un mode d'abreviation
qui finit par ne plus se distinguer d’un compose, il reflete certaines
notions pregnantes d'une culture et d'un culte herites. On n"a pas
assez tenu compte des enseignements que donnent a ce point de vue
les dvandvas avestiques, dont une liste (sujette a revision) a ete
dressee par Bartholomae, BB. x, pp. 267 sq. Et cependant il s'y
conserve autant et menie plus d'antiquitfe que dans ceux du vedique.
En ce qui concerne la forme, il est remarquable que les regies
formulees par Panini sur I'ordre des deux elements dans les dvandvas
vediques (Wackernagel, AUind. Gramm., ii, 1, § 70. pp. 165 sq.)
s'appliquent aussi en avestique. Quand les mots sont d’inegale
longueur, le plus court vient en tete*: aedrya aedrapaiti " eleve (et)
maitre ” ; dpa urraire " eaux (et) plantes " : xsvlhn dzuiti " aliments
liquides (et) solides ” ; tMusl atayditl " force (et) perseverance ” : sairi
rarazdrie confrerie (et) communaute ". 2° Quand les deux mots
sont parisyllabiques. celui qui commence par une voyelle vient
d'abord : aestm barasma “ bois (et) feuillage " ; aesma baoiSi " bois
(et) parfums ” : arazahi saiahi " ouest (et) est " : aspa vlra (dans le
compose aspa .vlra .gan-) chevaux (et) guerriers". 3“ Si les deux
elements sont parisyllabiques et d'initiale consonantique. la preseance
est accordee au theme en -i- ou en -u- : pdyd dworastdra " tuteur (et)
fabricateur ” ; pasu vlra " betes (et) hommes " ; fsaoni- vqdira-
(voir ci-des.sous). Les dvandvas vediques et avestiques sont done
soumis aux memes regies et reproduisent des modeles indo-iraniens.
Bien que le vocabulaire traditionnel ait ete de part et d'autre
rcnouvele, il subsiste dans la langue du rituel une correlation precieuse.
sur\dvance du culte commun : ved. idhmAbarhis- et av. aesma barasma
“■ bois (et) feuillage " (pour I'offrande). Le neutre correspondant a
V. barhis-, qui est av. barazis-. ayant pris le sens de " coussin " (cf.
Willman-Grabow.ska. Symb. gramm. Rozwadnwski . ii. pp. 167 sq.) a ete
remplace par le terme consacre barasman- “■ litiere de branchages ".
qui en est derive.
^ Sur cette question on lira avec fruit les penetrantes observations de
J. Wackernagel, Fesischr. Binz, Bale, 1935, pp. 33 sq.
406
E. BENVENISTE
3Iais par dela la communaute indo-iranienne, c’est a des representa-
tions indo-europeennes que certains dvandvas avestiques se referent.
Parmi les survivances propres a I’indo-iranien et a I'italo-celtique
CWackernagel, KZ. xliii, p. 295 ; Yendrves, MSL. xx, p. 280), on a
deja releve la concordance de umbr. dupursus peturpursus = vM.
dvipddag cdtuspadah " bipMes (et) quadruples Or la locution est
connue egalement de I'avestique, dans les couples bizangm- caOwara.
zangra- et bipaitistana- cadivara .paitiMana-, employes selon les etres
favorables ou nefastes ; le terme pad-, de connotation neutre, a ete
remplace par zangra- (resp. paitistana-) ” jambe ” qui permettait
d'opposer explicitement les creatures bonnes et mauvaises. Une
seconde correlation, connue elle aussi, est propre a I'av’estique seul
et a I'italique ; av. pasu inra = lat. pecadesque uirosque (Omde. Met. i,
286), undar. veiropeqm ; cf. encore lat. hominesque gregesque. Peut-
etre doit-on en retrouver la plus ancienne attestation en hittite, dan.s
(’expression antuhsas GUD UDU “ hommes, boeufs (et) moutons "
(par exemple Kleinasiat. Forsch. i, p. 168, et n. 8), a condition que les
ideogrammes GUD UDU recouxTent une designation unique du
gros et du menu betail.
Nous compterons une troisieme et une quatrieme correspondance
du meme ordre entre I'avestique et le latin. Le compose av. aspa.vlra.
gan- “ tueur de chevaux (et) de guerriers " suppose un dvandva aspa-
rlra- qui est aussi implique par les locutions aapanhdbo vlranhabo
(Aog. 78) et, avec une legere variante, at,po .garam nara .garam (Y. ix,
11). Les souverains Achemenides reproduisent le meme tour quand
ils vantent la terre d Iran “ aux beaux chevaux. aux beaux hommes ”
(huvaspa humnrtuja). Ce n'est pas un hasard si le latin dit, dans la
meme succession, eqnis inri.^(que). quoique I'e.xpression ait ete
specialisee dans la langue militaire pour designer I'eusemble des forces,
cavalerie et infanterie.
^ oici enfin le quatrieme dvandva conserve au.x deu.x extremites
du monde indo-europeen et propre au meme vocabulaire. A cote
de pasH-. il existe en av'e.stique une forme suffixee pasula- “ animal
domestique (cf. skr. pagidu-) qui fait couple avec daitila- “ bete
■sauvage : ^ t. xiii. 74, nrund . . . pastikanqm . . . dailikanqm ^
■‘les allies des animaux domestiqiies (et) .sauvagos ” ; opposition
attenuee par une re.striction nai’ve A", xxxix. ] sq. (cf. AY. xiii,
.Afire* rh.ique mot fie ce p.o,*.'cage, Ic texte portc i/rizrnnairlp. probablemcnt pour des
raison.* tie liturgie : rc-nonct- de chaquc mot de la priere appelait le repons yazamnide.
Autrement la suite grammaticale est intaete ; cf. trad. Lommel. p. 121, n. 1.
SUR QUELQUES DVANDVAS AVESTIQUES
407
154) uruno 'pasukanqmca yoi nd jijismtt “ les ames des animaux
domestiques qui nous nourrissent " ; daitikanqnwa aidyunqm . . .
uruno “ les ames des fauves utiles Ces deux mots ne s’emploient
pas Fun sans Fautre ; formes pareillement, ce sont des mots-rinies
qui peuvent, comme pasu vlra, etre dissocies par -ca. La locution
pasuka- daitika- fait penser immediatement aux ferae pecudes de
Lucrece i, 14, ou M. Ernout (BSL. xxiv, pp. 72 sq.) a reconnu un vieux
juxtapose “ animaux sauvages let) domestiques Comparer encore
lat. pecua belluasque (Naev. ap. Non. 159, 6). L’opposition des
deux especes d'animaux, pour naturelle qu'elle soit dans une societe
d'eleveurs de betail, ne se marque dans aucune autre langue, a notre
connaissance, par un dvandva semblable. Mais cette locution a ete
adaptee a des vocabulaires differents. Tandis que le latin conservait
dans fera Fancien adjectif *ghuer, *ghwero-, il y a ete substitue en
avestique un adjectif nouveau *data- datika- (probablement derive
de dant- “ dent ") qui reste encore en usage : phi. dot, pers. rfuS, sogd.
Sc *SoC, saka data. On notera en outre que, dans pasuka- daitika-,
la priorite du theme en -u- (cf. p>asu vira) est respectee.
A la meme categorie appartient encore Fexpression/sao«i- rqdwa-
dont le sens a ete defigure chez Bartholoniae par la traduction
“ Uppigkeit und Heerde " (Wb. 1027). Le nom vqdtea- designe
certainement le gros betail et sert normalement pour le troupeau de
bceufs ou de chevaux. metaphoriquenient pour une troupe d'hommes.
Si rqdwa- est souvent associe a fsaoni- et s'il denomme le troupeau
de gros betail, il est evident que fsaoni- doit signifier troupeau de
petit betail ”. En effet, on lit IV.. v. 26, apres une serie de locutions
formees de deux noms. uye fsaonlsca rqOivdca qui doit se traduire :
■■ a la fois le menu et le gros betail." De cet emploi ne saurait etre
disjoint celui de IV., i.x. 9 (ou Bartholoniae voit cette fois un adjectif
“feist, iippig ") : yada azoni fsaoni vqOira am.bardni “ puisse-je
apporter du menu et du gros betail " ; souhait forniule par Yama
qui vient de sacritier des chevaux, des bceufs et des moutons. Dans
le dvandva fsaoni vqdira se denonce ainsi une expression indo-
europeenne, la meme opposition des deux ti-pes de troupeaux qui est
rendue chez Homere par ttiuv et dyeXg. en latin par pccus et
armentum. Le sens cjue nous donnons k fsaoni- etait par avance a.ssure.
puisque la famille entiere de *peku se rapporte au mouton. a Fanimal
“ pourvu d'une toison ”. Le verbe/s«- (dont fsaoni- est derive comme
skr. yoni- de yii-) participe aus.si de cette signification, cox: fsuyant-.
on ne le remarque pas assez, est le nom de celui qui fait paitre le petit
408
E. BENVEXISTE —
betail. tout coinme fr. berger {*berbicarius). II faut done le distinguer
de vastrya- qui s'applique au pasteur des bceufs. Nous pouvons alors
proceder a I'analyse de Texpressioii vdstryo fsuyant- qui. dans la
hierarchie des classes sociales. qualifie le troisieine etat. celui des
agriculteurs. II est clair que vdstryo fsuyant- doit s'entendre comme
un dvandva et comprend deux termes distincts ; " pasteur du gros
et du petit betail." C'est une locution ou deux termes juxtaposes
ont fini par faire corps, exactement comme dans lat. patres conscripti
(= patres et conscripti).
A I'interieur meme de I'univers iranien. un dvandva fournit la
solution d’un probleme d'histoire religieuse. On connait, dans le
groupe des Amrta Spantas, les deux genies Harvatat et Amrtatat
(generalement ecrits Haurvatdt et Am?r3tdt) Integrite et Immortalite
qui veillent respectivement sur les eaux et sur les plantes. On s'est
souvent demande d'oii ils proviennent et pour quelle raison ils sont
constamment associes. Certains ont meme voulu les prendre pour la
figuration iranienne des Alvins. II ne faut pas chercher si loin. Le
groupe entier des Amrta Spant.as est tres probablement la transposition
abstraite des anciens genies des elements, dont la fonction. sous leur
designation nouvelle, n‘a pas varie. C'hacune de ces entites est attachee
a un aspect du monde materiel, terre, metaux, eaux, etc., et le
symbolise. Or la plus ancienne mention du couple Harvatat-Amrtatat
en illustre la veritable natme. C’est le passage gathique Y. xxxii. 5.
td dabmaotd imsim hujydtois amsrotdtascd par ce (mefait), vous avez
ftustre rhomme de la bonne me et de I'immortalite ”. L'abstrait
hujydti- ‘'bonne me” est I'equivalent theologique de haiirvatdt-
“ integrite (physique), bonne .sante A notre avis, il faut entendre
I’expression hujydtois amdrotdtascd dans une acception materielle :
■'vous avez frustre I'homme de I’eau et des plantes" (cf. li, 7.
apased urvarased au.jratdtd haurvdtd). Dans toute cette Gatha, le
reformateur fait allusion a une serie de crimes bien specifies : con-
sommation de la chair du bceuf, injures au boeuf et au soleil, destruction
de paturages. violences sur des Zoroastriens, etc. II accuse expresse-
ment ses ennemis d’avoir ‘‘ detruit la vie” (11 morondm jyotum).
d'avoir ‘‘detruit la vie du boeuf” {gms mormdon . . . jydlum). Ces
mefaits ont pour consequence naturelle d’oter a I'homme ses moyens
d'existence. Nous ne croyons pas forcer le sens du vers, mais bien
au contraire I’accorder avec le reste de la predication, en lui donnant
son plein sens concret. Par ailleurs, dans I’Avesta recent, la liaison
de Harvatat-Amrtatat avec les eaux et les plantes fait I’objet d'une
SUE QUELQUES DVANDVAS AVESTIQUES
409
tradition claire et constante. II devient alors an moins vraisemblable
que le couple Harvatat-Amrtatat est la projection mjdhique du
dvandva djpa urvaire “ eaux et plantes atteste maintes fois (cf.
Bartholomae, Wb. p. 327) et indirectement par les expressions
tacat.dfa uxsyat.urvara (Y. xvi, 8); tai.dpd uxsyat.urvaro (Yt xiii,
43, cf. X, 61). Les deux termes se suivent dans I’ordre fixe par la regie I
(p. 405), aussi bien Harvatdt-Amrtaidt que dpa urvaire. En lui-meme
le dvandva dpa urvaire, chez un peuple d'agriculteurs, n'a pas besoin
de justification ; c'est le dvandva des noms di\'ins qui en demandait
une. Le proces mytbique que nous restituons acheve de montrer dans
ces entites des elements divinises.
VOI.. Vlir. PARTS 2 ASD 3.
La Charrue vedique
Par Jules Bloch
I L y a eu tin verbe indo-iranien pour designer le labour : RV. imp.
karsa, krsalu, av. opt. pairi-karsdit ; il s'y rattacbe des substantifs
designant le sUlon : av. karsa-, karsi- f. ; RV. sans doute krsi-, SBr.
karsu-.
Par centre les noms Sanskrits de la charrue n’ont pas de corres-
pondants connus dans I’lran.
L’un de ces noms, qui vit encore aujourd’hui, skr. hala-, n’est pas
atteste avant Panini. Est-d cependant plus ancien que le Sanskrit ?
On en a propose plusieirrs etymologies indo-europeennes, dont
malbeureusement aucune ne s’impose (v. Walde-Pokorny, Vergl.
Tr6. der Indog. Spr., i, p. 629 ; Scheftelowitz, ZIL, ii, p. 278). Le
Rgveda offre deux autres noms encore :
L’un, lihgalam, garde encore son sens ancien, non seulement en
indo-aryen (mar. ndgar, etc., J. Bloch, Marathe, p. 357), mais dans
d’autres families indiennes : telugu ndgal, canara negal, tamoul
ndnjil (et brahui langar) ; santal nakel, mundari naeal ; il semble
bien y avoir ete emprunte des Aryens en meme temps que I’instrument,
ou du moins la forme perfectionnee de I’instrument ; car si Ton suit
M. Przyluski (BSL., xxiv, pp. 118-23) on verra dans Jdfigala- im nom
indigene prehistorique ayant d’abord designe le baton a creuser des
trous et ayant peut-etre conserve cette valeur tres tard, a en juger
par la citation du commentaire a Nirukta vi, 26 ; bhumim
bhumisaydms caiva hanti kdsiham ayomukham iti abhidkeyaprasangdt
Idngalam aka. Dans le Veda, le Idngala- est en effet caracterise comme
pourvu d'une pointe metallique, pavira-. Le soc a d’ailleurs une
designation, phdlah, qu’on rapproche d’un nom persan de la
charrue, supdr.
L’autre nom vedique, sfrani (variante Kap. S. sila-), est dcja rare
en pali, et ne semble plus usuel a I’epoque moderne que dans le sens
derive de champ non afferme et cultive par son proprietaire (Turner,
Nep. Diet., s.v. sir ^). Rien ne caracterise cet instrument dans le
Veda, si ce n’est la mention d'attelages nombreux : AV., viii, 9, 16,
sadyogdm stram “ charrue attelee a six ” — dans une enumeration
de choses allant par six ; TS., i, 8, 7, 1, dvddasagaidm strain ddksind
412
JULES BLOCH —
“ le salaire (du rite decrit sera) une charrue a douze boeufs Mais
TS., V, 6, 21, 1, deux moutons le trainent, siravahau dvl : ce n’est
done pas necessairement un instrument lourd, contrairement a
I’indication de Macdonell- Keith, Vedic Index (recueil cependant
precieux, qui nous a servi de guide en tout ceci) ; il n’y a du reste
sans doute rien de precis a tirer de ces chiffres, dont le premier depend
du contexte et les autres designent peut-etre des objets figures.
Voila done au moins deux noms pour le meme instrument. Mais
est-ce bien le meme instrument ? Meme sans tirer parti du texte
relativement tardif cite plus haut, on remarquera que la phraseologie
du Veda ne traite pas les deux mots de la meme fajon. L’unique
Idhgalam de RY., iv, 57, 4 = AV., iii, 17, 6, s'accompagne du verbe
attendu Jears-, le meme verbe s'appliquant du reste aux betes de trait
(non specifiees) et aux conducteurs ;
sundm vahdh sundm ndrah
sundm Irsatu Idhgalam.
C’est naturellement le meme verbe qui designe Taction du soc :
X, 117, 7, krsdnn it pMlah
iv, 57, 8, phdla vi hr.santu hhimim.
Jlais avec stra-, les deux fois qu'on le trouve dans le Rgveda
(x, 101, 3-4 = AV., iii, 17, 1-2 ; cf. TS., iv, 2, 5), c’est ynj- qui
Taccompagne :
3. yundkta sfrd vi tjitgd tanudhvam (AV., TS., tanota)
krte yonau vapafehd bijam . . .
4. strd yunjavti kavdyo
yugd VI tanvate p'rOiak.
De meme, avec lahgala- ; prahliid- TS., au, 6, 7, 4, pratan- Kaus.,
XX, 1, et nulle part semble-t-il yuj- ; mais yuj- avec stra- deux fois
SBr., KSS., xvii, 27, ApSS., .xvi, 185 ; karp seulement avec le compose
svraldhgalam, dont il sera question plus bas (je dois ces confirmations
a 31. Renou).
Est-ce forcer Tindication que de remarquer que dans le Rgveda
on trouve mentionne a cote du stra- le joug, qui n’est qu'une piece
de la charrue, et meme une piece detachable ? C'est ce qui a conduit
Grassmann (non suivi par AVhitney) a traduire “ attachez les cordes ”,
spannt an die Strange (mats iv, 57, 4, .sundm varatrd badhyantdm).
Il a du etre guide, non seulement par le parallelisme avec yugd et le
sens ordinaire de yuj-, mais aussi par la possibilite de rattacher
etymologiquement stra- a la racine de skr. aor. dsdt, opt. simdhi, pr&.
LA CHARRUE VEDIQUE
413
syati, part, sitd- ; av. pf. kisdyd, part, hita-, inf. dhoiOoi, dont le sens
fondamental est “ attacher ” ; sur Vi dans ce type d’alternances, cf.
EV. Tcsitd-, AV. ksind- et Wackernagel, Altind. Gr., i, p. 87 s. Dans
cette interpretation, stra- correspondrait exactement a v.h.a. et all.
seil, v.sl. silo (Walde-Pokorny, ii, pp. 463-4). Les strd seraient
alors sans doute les lanieres qui joignent le joug a la charrue ; I’auteur
du Bihar Peasant Life, a qui une etude comme celle-ci se dedie tout
naturellement, en a donne les noms modernes, § 18 : ndran ou Idran,
nddha ou Iddha (a vrai dire aucun de ces noms ne rappelle stra-).
Et puisqu’il faut une autre laniere pour retenir le noeud fait par celle
dont nous venons de parler (v. Grignard, Oraon-English Diet., s.v.
ugtd ; Hoffmann, Encycl. Mundarica, s.v. cutu-nahgali), le pdrisiryam
de SBr., vii, 2, 2, 3, qui est fait d’herbe muSja tressee en natte triple,
serait alors cette seconde laniere.
Mais dans ce cas, comment expbquer que le strain puisse etre de
bois ? Or e’est ce que dit le meme passage du ^atapatba Brahmana :
strain yunakti . . . audmnharam bhavati. S’agirait-il alors d'une cheville
de bois ? Justement sail est d'apres H. M. Eliot et Beames, Memoirs
on the jV.PF. Provinces of India, ii, p. 342 et fig., le nom de deux des
quatre cbevilles qui retiennent ensemble les deux barres dont le joug
est fait ; les sail sont les cbevilles exterieures, les gala les cbevilles
interieures, plus longues ; et il y a une corde {pdrisiryam ?) pour
rejoindre un sail et un gata sous le cou de la bete.
Mais ici se presente d’abord une difficulte bnguistique. Sans doute
*saila- est le derive a vrddbi normal de sila- : mais ceci n’est vrai
qu’en Sanskrit. Dans une langue moderne, on s’attend qu'une
dipbtbongue ai provienne de deux syllabes differentes du Sanskrit ;
la forme attendue est celle qu'on trouve dans bib. sel, sell {Bihar
Peasant Life, § 109), b. sell “ collier de fil noir des ascetes ”, s. selhi
“ corde faite de cbeveux ”, cf. Turner, Nep. Diet., s.v. sell. D'autre
part sail a un doublet plus developpe, que nous connaissons grace
encore au Bihar Peasant Life : § 15, “ Tbe outer pins, wbicb join tbe
two bars of tbe yoke are saild or samail to the west and kanail to tbe
east ” ; § 16, “ tbe inner pins are samail or (Patna . . .) samaila and
(Gaya) samaiya.” Si bien que nialgre qu'au Tirbut NE. samail ou
samel designe la corde passee au cou de la bete (§ 18), on pent
soup 9 onner dans ces mots des composes dont le second terme serait le
nom de la cheville, skr. klla- ; cf. le synonyme de kanail qui est a
Bbagalpur kan-killi. La cheville ajustant le manche au corps de la
charrue est taraila, avec la meme finale.
414
JULES BLOCH
Remarquons enfin que dans I’liypotliese “ cheville ” comme dans
I’hypothese “ corde ”, on ne s’explique pas ce que pouvaient etre les
svra- a six ou douze betes.
On est done amene a cbercher ailleurs et le sens et I’origine de sira-.
Qui observe les langues non aryennes de I’lnde sera d’abord frappe
de la coincidence avec un nom dravidien de la charrue : gondi ser,
kui seru, tel. can. tarn. er{u), coexistant avec nagal, etc.^ Malheureuse-
ment, meme si les probabilites n'allaient pas centre I’idee d'un emprunt
fait par I'aryen, ce que nous savons jusqu’a present du phonetisme des
deux groupes n'explique pas non plus le passage de e a I : tant qu'a
rapprocher les deux series, on songerait plutot ici encore a I’entree
en dravidien d’un derive a vrddhi.
Mais en indo-aryen meme il est permis de faire etat d’une racine
indo-europeenne, partiellement homonyme de celle qui a ete examinee,
a savoir celle de *se- “ semer ”. Que cette racine ait pousse des rejetons
jusqu’en Sanskrit se reconnait an nom du “ sillon ”, ou plutot, puisqu’il
s’agit ici d’un adjectif verbal au feminin, de la “ terre ensemencee
sUa : la place de I’accent s’expliquerait par im changement de fonction,
comme dans d’autres mots que M. Renou me signale amicalement :
dsta-, maria-, surta, sans doute vrdta-. La signification primitive du
mot ^ parait encore dans les derives comme Pap. Am. sityam “ champ
laboure ” (plus tard “ ble, grain ” ; cf. khovar siri “ orge ”, oil r
provient de t) ; elle subsiste encore abondamment, v. Turner, Nep.
Did., s.v. siyo (ajouter bhadr. sith f. ; sithnu “ faire un second labour ”).*
Du nom du sillon se rapproche aisement celui de la raie des cheveux,
et par suite du sommet de la tete ; et d’autre part, suivant une
evolution bien constatee (Vendryes, Melanges P. Boyer, p. 13 s.), de
la limite ; skr. simdn- m. (et simanta-), d’ou pj. s%, etc., v. Turner,
Nep. Did., s.v. simdna.
^ En santal et inundari, si- signifie “ labourer ”, mais er-, her- “ semer ” ; d’autre
part en sora or- " labourer ”, ertub- “ charrue ” en regard de lud- “ semer La suite
permettra d’lnterpreter ees eciianges. Et il ne faut pas oublier que le vocabulaire n’a
pas plus de raison d'etre indigene ici que I'instrument.
^ On sait la fortune qu'il a eue comme nom propre. Est-il permis de se demander
si en face de Slid, la rerre labouree, Ahalyd^ trait reusement seduite par Indra, n'a
pas ete d’abord la terre interdite au labour ?
Le lien eritre siram et slid a-t-il ete senti ? Dans ie Jataka do Mend{h)aka, il
est dit de son csclave que quand il laboure avec une charrue, sept sillons se ferment :
ekanangnlena kasantassa satta sltdyo gacchanti (Mahavagga, vi, 34) ; ce que le
Divyavadana, p. 124, 1, 7, exprime ainsi : sa yadaikarn halaslrani krsatiy tadd sapta
slrdh kr^td bhavanti. Faut-il corriger le texte, ou admettre un nouveau stra- signifiant
sillon ’ , d’oii deriveraient mar. sera, etc. ? Chez Amara slrah se situe entre goddranam
et samyd.
LA CHAKRUE VEDIQUE
415
II est remarquable qu’il existe en marathe des mots de type *saira~
et de sens voisin : si Ton pent rattacher sera “ bout ” a sird atteste
aussi en Hindi avec le sens de “ limite de village ”, suffira-t-il pour
I’expliquer d’invoquer skr. sirah “ tete ” ? En tout cas ce recours
serait insuffisant pour sen “ ruelle, passage entre des clotures
Mais ici une objection grave se presente ; semer n’est pas labourer ;
nulle part ailleurs i.e. *se- ne designe le travail de la charrue, ni *ar- ^
les semailles, pour lesquelles existe en Sanskrit une racine vap- (indo-
iranienne ? v. Morgenstierne, AO., i, p. 256), laquelle est encore en
usage, V. Turner, Nep. Diet., p. 645 s.v. uhdunu. En fait, il arrive a
vap- d’accompagner le meme contexte que hirs- en deux passages
du Egveda :
i, 117, 21, yavarn vfkendsvind vdpantd
viii, 22, 6, yavarn vrkena karsathah.
Mais on ne pent rien en deduire, meme si ce “ loup ” avec lequel
les Aivin tour a tour sement et labourent devait finalement etre
un veritable nom d’instrumeut, dont le nom appartiendrait a la racine
de lit. velku “ je tire ”, laconien evXana “ charrue ” (Vendryes, ibid.,
p. 14 ; Walde-Pokorny, i, p. 308). Y avait-il done dans I’lnde ancienne
une charrue susceptible de semer ?
En tout cas elle existe dans ITnde moderns. Consultons a nouveau
le Bihar Peasant Life : au § 857 sont decrites trois methodes de semailles ;
a la voice, le long du sUlon a la suite de la charrue, enfin avec le idr
ou tar, instrument decrit a son tour au § 24, sous le nom de drill-
plough, charrue-semoir. La charrue ordinaire s’appelle har ou hal,
et a Gaya Idiigal.
Quant au tor, e’est une charrue a laquelle s'ajuste un tuyau,
hdsd ou edgd, surmonte d’un receptacle appele mala, mdleva ou jmila
ou encore ukhri, akri, ou on verse la semence. Le Hindi sabdasagar
de Benares confirms cette description precisement aux mots bdsd
^ Risquons encore une hypothese. Si *se~ a des rejetons dans I’lnde, est-il
impossible de reconnaitre *ar- dans Tindo-iranien ttnard, qui serait alors egalement
un adjectif feminin designant la terre vastement {uru-) labouree, ou celle dont on
laboure les sillons cf. lat. uruos, gr. ovpoy, etc. ; pour I’accent sur le 2© terme
du compose, v. Wackemagel, ii, § 114d, cf. § 96 6, 8) ? L’unique 'phdrvara-y artificiel
ou non (v. Henry, MSL.y xiii, p. 172 ; Oldenberg, Rgveda Soten ad x 106. 2) en serait
une imitation, construite soit sur phdla-, soit sur le radical des mots qu’on trouve
dans le meme hymne, pharpiharaX, parphar’ika-, qui est peut-etre celui de gr. ^dpos
“ charrue
416
JULES BLOCH
et aJcriy De meme chez EUiot-Beames, p. 340, hal ou har est une
charrue “ si Ton peut donner un nom aussi noble a un instrument sans
centre ni oreillon ” ; mais p. 227 le Msd (a Delhi orna) est le tuyau
par on la semence descend dans la machine a semer ; nous aliens
voir plus has ce qn’est sans doute cette machine ; on aioute qu'au
Nord-Ouest le bdsa est generalement attache a la charrue proprement
dite — comme au Bihar.
Comme au Rajputana aussi : en 1809 Th. D. Broughton rencontre
a Udaipur Finstrument suivant {Letters written in a Mahratta Camjp,
p. 215) : “ They use a drill-plough . . . The drill part consists of a piece
of bamboo, about two feet long, split and widened at the end, where
it is covered by leather, so as to resemble a funnel, and which is fixed
behind the main stick of the plough. In this wide part the driver
keeps his left hand, filled with grain, with which he at the same time
steadies the plough ; while in the right hand he holds the reins
of rope and a long stick, with which he guides the bullocks that draw
it. One man only is required for each plough.”
Et au Deccan : en 1800 Fr. Buchanan (A Journey from Madras,
i, 283, et fig.) voit a Kolar semer le sorgho “ by means of an instrument
named sudiky, which is tied to the handle of the plough Get
instrument peut se compliquer : le Manual of Administration of the
Madras Presidency, iii (1893), decrit aussi le gorru telugu, s.v. gorroo :
“ A seed drill. It consists of a beam with three or six shares fitted into
it, and a corresponding number of bamboos attached to a zaddigam,
or drill, into which the seed is thrown and is thence carried into the
ground by bamboos. . . . The gorroo is also used as a plough.” Le
curigy avec lequel Buchanan a vu a Banawasi semer certaine espece
de riz comporte quatre tubes (iii, p. 236 et fig.) ; mais ailleurs et pour
d’autres graines on en emploie qui ont douze tubes (de meme R. H.
Elliot, Experiences of a Planter, 1871, ii, p. 40) ; et il arrive que derriere
les douze tubes du curigy qui versent le ragi on attache Funique sudiky
qui sert a ajouter la graine de avaray ou tomary (i, p. 285 et fig. ; cf.
i, p. 377).
Ces instruments, qui realisent couramment le miracle attribue a
Fesclave Mend(h)aka par la legende bouddhique (v. p. 414, n. 3) sont,
non pas des derives, mais des parents tres eloignes des semoirs
europeens, qui ne datent que du xvii® ou du xviii® siecle (v. Encycl.
' >Iai3 je n’y trouve pas vaima, uirva donnes par Fallon et Platts ; mots derives
de skr. Mr- ; cf. mar. per- “ semer ” de prakir-.
LA CHAKRUE VEDIQUE
417
Britannica, 14® ed. s.v. sowing ; Leser, Entstehung und Verbreitung des
Pjluges, Anthropos Bibliothek, 1931, p. 453), et qui sont independants
de la charrue.
En voila assez sans doute pour marquer I’extension de la charrue-
semoir dans I’lnde. Quant aux mots qui la designent, ou bien ils ne
sont pas clairs, ou bien ce sont des termes generaux designant sa
matiere ou sa forme : hasd derive de bds “ bambou ”, cbgd designe
un receptacle ou un tube de bambou (v. Turner s.v. cugo), brabui
ndri veut dire “ tube ”. Qu’y aurait-il d’etrange a ce que stram de
son cote eut exactement la valeur du fran§ais “ semoir ”, entendant
par la soit I'ensemble charrue plus entonnoir, soit plus simplement
I'entonnoir lui-meme ? L’expression strayn yuj- signifierait alors soit
“ atteler la charrue-semoir ”, soit plutot “ ajuster I'entonnoir a la
charrue ” suivant I’operation plusieurs fois mentionnee ci-dessus :
et dans ce cas siraldhgalani Vadh. S. {AO., vi, p. 117, avec prayacch- ;
p. 237, avec saynsrj-) designerait I’assemblage, et SBr. pdrisiryam la
corde (dessinee par Fr. Buchanan) maintenant cet assemblage.
Pour donner force a I’interpretation et a I'etjunologie ici proposees
de stram, il faudrait prouver I'existence de la charrue semeuse dans
rinde antique. La demonstration directe est impossible. C’est encore
une chance que les bas reliefs de Bodh Gaya et du Gandhara nous
renseignent sur les types de charrue proprement dite en usage aux
alentours de I'epoque chretienne et nous assurent de I'antiquit^ des
types actuels. M. Leser, qui en a reproduit trois exemplaires dans
I’ouvrage capital auquel nous nous sommes deja refere {Entstehung . . .,
planches 17 et 18, cf. p. 382), a montre que leur extension depasse
ITnde et remonte a une epoque tres haute. II a de plus mis en valeur
(dans le meme ouvrage, p. 245. et dans son article de la Festschrift . . .
P. TT. Schmidt intitule “ Westosthche Landwirtschaft ”, pp. 416-19,
avec figures) le fait que I’entonnoir a semences egalement se retrouve
en Syrie et en Chine, et des une epoque tres ancienne en Mesopotamie —
mais non en Egj-pte. II en reproduit (respectivement, p. 244 et
p. 417) une image datant du troisieme millenaire et une datant
du second millenaire avant J.-C. ; il en signale d’autres, par exemple
celle du palais de Sargon, celle du monument d’Asarhaddon du vii®
siecle ; le tube a semences apparait du reste ici comme encastre
entre les deux manches de la charrue. En voila plus qu’il n en laut
pour assurer de I’antiquite reculee du precede, et cela a I’interieur d une
aire ou M. Leser montre que d'autres techniques se retrouvent
418
LA CHAERUE VEDIQUE
repandues, et conlinees : par exemple, la herse simple ou le rouleau
servant de herse (la herse aussi manque en Egypte). Cette com-
munaute, et ce qu’on sait ou devine par ailleurs des relations
prehistoriques entre la Mesopotamie et I’lnde, engage a admettre que
la charrue semeuse est dans I’lnde bien anterieure a la compilation du
Veda, et que par consequent il ne serait pas etonnant de I’y trouver
mentionnee.
The Dialectical Position of the Niya Prakrit
By T. Burrow
rpHE “ North-Western Prakrit ” as Konow has called it is repre-
sented by the following documents.
(1) The two versions of Asoka’s edicts preserved at Mansehra’
and Shahbazgarhi. At this stage many of the characteristic features
of the language have not yet developed, e.g. sr > s, sv> sp.
(2) The later Kharosthi inscriptions, mostlv^ short, collected by
Konow in the second volume of the Corp. Inscr. Ind.
(3) The Kharosthi manuscript of the Dhammapada discovered
near Khotan {Manuscript Dutreuil du Rkins).
(4) The Kharosthi documents from Niya, representing the adminis-
trative language of the Shan-Shan kingdom in the third century a.d.
In the Journ. As., 1912, pp. 337 ff., J. Bloch examined the
dialectical peculiarities of the Manuscript Dutreuil du Rhins and
showed that they appeared in modern times in the languages of the
North-West.
The later Kharosthi inscriptions of the North-West along with the
Kharosthi Dhammapada, were taken by Konow as a basis for a detailed
treatment of the North-Western Prakrit in his introduction to the
second volume of the Corp. Inscr. Ind. He did not bring in the Niya
documents, although, owing to their extensiveness and comparative
freedom from literary influences, they are the most important of all.
It is the purpose of this paper, not to give an account of “ Niya ”
phonology, but to discuss specifically those points in which it coincides
with or differs from the several varieties of the language of the North-
West mentioned above ; and on the basis of this to show that out of the
modern languages of the North-West Torwah is the one which shows
the most striking resemblances to the dialect of Niya.
I. Comparison with the Kharosthi Versions of Asoka
Points in common are : — -
(1) Preservation of the distinction between the three sibilants
s, s, and s, which was early obhterated except in the North-West.
(2) A better preservation of conjunct consonants, especially groups
with s and r (st, tr, rt, etc.).
(3) Development of the vowel r into n or ru. In the rest of India
420
T. BURROW
there is usually no trace of an r. Examples from Asoka are : (Mansehra)
mrige, (Shahb) mrugo, (Mans, 5) vudhresu (but Sh. vudhesu), (Sb. 11)
graha (M. geha), Sh. 13 dridha. In the Niya Prakrit usually ri (which
may be written r in imitation of Sanskrit), e.g. etrisa “ such ”, krita,
krisati, glirida, dritha, triti, also krta grha, etc., which is merely a matter
of spelling.
(4) Assimilation of s to s in the words anukasanam, anusakisainti
(S. M. 4).
In Niya kakana (510).
(5) sg> k S. M. manuka, anukakikamti, etc. ; Niya manuka,
karikati, etc. The writing sy which occurs in Niya side by side with
k is, of course, merely due to the influence of Sanskrit.
(6) The primary endings are appended to the optative ; Aioka :
patipajeyati, apakareyati, nivateyati ; Niya : anuvarteyati, avarajeyamti
{apa-radh-), visarjeyasi, etc., etc. These forms occur in Pah, usually
in the later texts (Geiger, Pali Gr., § 127). In Prakrit according to
Pischel (§ 459, § 462) we do not find the primary endings in the third
singular, although they are common enough in the second singular,
which suggests that the process started in the second singular as an
effort to distinguish the endings which had become confused, and was
not in all dialects extended to cover all the persons, but only con-
sistently in the North-West and in that dialect, presumably western,
which lies at the basis of Pali. The -d-, which is always long in these
forms, must come originally from forms in -yat, kurydt, etc., which also
explains Prakrit forms like vattejjd, beside vattejja and vatte (Pischel,
§459).
(7) Indeclinable participles in -ti. Asoka : tithiti, draketi, vijiniti,
aloceti ; Niya ; kruniti 341, apruchiti 39, vajiti “ having read ” 376, etc.
(8) Infinitives in -anaye. kxamanaye S. 10 (other versions have
-lave). Niya : karamnae, deyamnae, etc., etc.
In some points the Niya Prakrit approximates more closely to one
of the Northern versions of Asoka than the other. Thus the change
from -j- to -y- (Niya •tnaharaya, etc.) occurs only in Shahb., e.g.
satnaya (= samaja-), raya, kaniboya. Similarly the assimilation of
dv to b. Shahb. badaka. Mans, duvadaka, Niya badaka. Also the changes
nj > « and ny>n; Shahb. vamnanato (Mans, probably viyarnj-),
Shahb. pmtia, hiratia (Mans, puna as in the East).
On the other hand Mans, agrees with Niya in turning tm. into
tv in atva (S. ata). Most important is the treatment of final -as of the
nom. sing., etc. In Shahb. it predominantly appears as -o, in Mans.
THE DIALECTICAL POSITION OF THE NIYA PRAKRIT
421
as -e. The Niya dialect agrees with Mans., e.g. tade and the usual
ablative termination -ade. This -e does not, however, appear in the
nominative singular because that has been confused with the
accusative and both appear as -a, but its original presence there is
attested by verbal forms like gademi < gato’smi, etc.
In addition there are numerous points in which the Prakrit of
Niya is differentiated from both the Kharosthi versions of Asoka.
Of course, there are the characteristics of a later stage of the language,
such as the weakening of internal consonants, but these are of no
importance from the point of view of dialectical distinction. Points
showing parallel different development are : —
(1) r appears as ri (r) much more consistently in Niya than in
Asoka. In the former ri is almost universal, while in the latter the
proportion of forms in which the r has disappeared is quite large.
Thus in Asoka, side by side with forms hke mrige, mnigo, graha,
vudhresu, we find a great number of forms like Jcita, kata, viyapata,
viyaputa, nivutiya, icsatena, bhataka, muto, vuta, edisani, paripucha ;
in fact the r is lost in the majority of forms, with invariable cerebraliza-
tion of a following dental. Also even when the r is preserved sometimes
we find cerebralization occurring — viyaprata, katra, kitra.
(2) A closely related point to the above, the combinations or
r + dental are better preserved in the dialect of Niya than in Asoka.
Cases of assimilation (with or without cerebrahzation) are the
exception in Niya. We find bhataraga “ master obviously a loan-
word, parivat “ to e.xchange ”, possibly so, katavo besides kartavo.
(Here there is no reason to assume external influence, but the verb
kar- is subject to irregularities because perhaps of its being used as
a kind of auxiliary and weakly stressed.) The same reason no doubt
accounts for the disappearance of -r- in the preposition mdha “ with “
(= sdrdham). But in the vast majority of instances the r is preserved,
e.g. kirti, vardhati, artha, ardJia, anuvarleyati.
In Asoka, on the other hand, the assimilation of r a following
dental is almost the rule, e.g. Shahb. anuvatatu (M. anuvatatu),
kaiava, anuvatamti, nivateti, etc. ; atha, nirathiya, vadhisati, vadheti,
vadhita, diadha. Bsually, as seen from these examples, cerebrahzation
appears, but not always. In Shahb. the only example of the
preservation of such a combination seems to be athra in iv, 10. There
are also a few examples where r is preserved along with cerebralization
of the dental, e.g. kitri and athra. In Mans., while assimilation is
common, we do find more examples of preservation than in Shahb.,
422
T. BURROW
e.g. nirathriya, athra, vadhrite, and vadhrayisati (these forms are
presumably just a way of writing artha, etc., cf. Hultzsch, Corp.
Inscr. Ind., i, p. Ixxxrvni), but even here assimilation is the commoner.
(3) The group -rs- is preserved in Niya varsa, darsida “ packed ”.
It is usually assimilated in Asoka vasa, hasamti, pasamda (also preserved
in prasamda with transposition of the r).
(4) -sc- appears as -c- in Niya, i.e. it is distinguished from c. In
Asoka no distinction is made.
(5) ts is preserved in Niya, vatsa, samvatsara. It is assimilated to
-s- (= ss) in Asoka cikisa.
(6) ms develops into mts in Niya samtsdra, tndmtsa. The nasal
is simply omitted in Asoka vihisa (= vihimsd). Similarly rns > me
in Niya samcaya (= samsaya), but Asoka sasayika.
(7) -ly- is preserved in Niya, e.g. kalyana. In Asoka it is assimilated
to y {kayana Mans.) and I (kalana Shahb.).
(8) -Ip- is asimilated in Asoka apa, kapa. It is regularly pre.served
in Niya alpa, silpiga “ artisan
(9) sv appears as sp in both versions of Asoka spagra, spamikena,
spasuna. In Niya it becomes sv, svasu, svasli.
(10) In Asoka -sm- of the locative sing, becomes sp, i.e. -aspi
(through *-asvi), or alternatively it is assimilated to -asi, uthanasi,
etc. Niya, when it does not use the old -e, has invariably -ammi.
Thus dialectically there are considerable differences between the
Niya Prakrit and the North-Western versions of Asoka. And the
interesting thing is that phonologically the language of Niya presents
a pronouncedly more archaic aspect than Asoka, namely in preser\'ing
better the consonant combinations such as -rt-, rdh, -rs-, -Ip-, -ly-, -sc-.
Nor is there any reason to think that this is a question of orthography
in Niya due to the influence of Sanskrit, because •when a.ssimilation
did take place there it was usually written, e.g. .st > th, dritha, etc.
Sanskrit forms do, of course, occur throughout the documents but
they are sporadic and tend to occur in those parts where the style
was slightly more elevated, for instance, in introductory formulae
to letters.
Obviously we cannot derive the Niya Prakrit from the language
of Asoka, and the most natural conclusion to draw from the fact
that phonetically it is better preserved is that its home is to be sought
further to the west. Because it seems clear (then as now) that the
more remote a language was in the direction of the North-West the
less liable it was to phonetic decay.
THE DIALECTICAL POSITION OF THE NIYA PRAKRIT
423
II. The Later Kharosthi Inscriptions of N.W. India
The inscriptions later than Asoka are not very extensive and are
spread over a considerable number of centuries. Consequently they
provide no satisfactory basis for determining the dialect of a particular
area at a particular time. Nevertheless, they furnish nseful points for
comparison on the one hand with the earher Asokan texts, and on the
other hand with the dialect of Niya. They more closely resemble the
latter than the former, naturally in being further developed (e.g.
in the weakening of internal consonants), but also in most of the points
mentioned above.
Thus we find that most of the inscriptions agree with Niya in
preserving conjunct consonants better than Asoka, e.g. Kartiya,
anugrakarthae, Khardaa, sardha, °mrdkia, samvardhaka, etc. rs,
on the other hand, usually appears assimilated in vasa, whereas Niya
has varsa.
Likewise (agreeing with Niya) there is no tendency to transpose
r, as is found in Asoka dhratna, etc. We always find dharma, °sarma,
karma, etc. This is in a way surprising because in the modern languages
of the North-West this transposition is general, and it is naturally to
connect it with the similar phenomenon in Asoka and assume that it
started very early. We do, on the other hand, find examples in this
period in the Eiharosthi Dhammapada, so that presumably it was
current in some areas of the North-West at this time and in others not.
Amongst other features may be mentioned ; the transition of
-nj- and ny to n, e.g. puna, nati ; the preservation of ks (however
it or the Niya cK were pronounced) as a separate soimd ; preservation
of the group st (medially but not initially according to Konow, p. cxi ;
the state of affairs in Niya seems to be roughly the same, but we do
find initial st- in stasyati alongside forms like thavamnae and thana).
st is assimilated to th as in all the varieties of the North-Western
Prakrit.
Likewise, just as in Niya we find samtsare, or perhaps samtsare
if Konow’s contention {BSOS., VI, 405) be right. Further, ir > s in
samana, savaa, sadha. Cf. Niya masu < smasru, sayati < srayate (with
the sense of “ seizes ”), samana, etc. The change was universal in the
North-West. The Kharosthi Dhammapada shows it and also a similar
treatment of sr in anavasutacitasa — anavasruta°. An intermediate
stage is represented by KLar. Dh.p. visravatena = visravatd. A
similar tendency to this latter is manifested also by the Sanskrit
of the Divyavadana, e.g. p. 450 Idldsrotasa, p. 553 srastd. Likewise
424
T. BURROW
in loanwords in Central Asia Saka ssamana, Sogd. hnn, Toch. samam,
and probably from Central Asia, N.Pers. saimn “ an idolater
In agreement with the Niya Prakrit -to- is assimilated to -p-
(=-pp-) in sapana (sattvdndm), ekacaparisai “41”; Niya caparisa
“ 40 The word dtman- is treated in a variety of ways. On the one
hand we find it developing to -tv- and through that to -p-, e.g.
apana-ge, atvan°, atvanasa ; Niya, compare apane (139) = dtnianah
and apanasya 201 (apanasya kritaga “ your own deeds "). In other
inscriptions we find as in Shahb. ata.
kv usually develops into kp — vikpasu, ikparaka. Similarly in Niya
akpa, kpiedaga “ white
-*asmi of the loc. sing, usually, as in Niya, appears as -ami {-asi
is doubtful in the list of forms given by Konow, p. cxi). There is no
trace of the Asokan -aspi.
sv is always preserved in these inscriptions — svakiya, svami, svarga.
There is no trace, either of the Asokan tendency to change it into
sp or of the tendency of the Niya Prakrit to produce kv.
The weakening of internal consonants has proceeded much along
the same lines as in Niya. It was customary to indicate this by
appending the /- sign to a letter, however it may have been pronounced ;
thus k(r) g{r) corresponding to Niya g, s{r) corresponding to Niya s.
There is a similar tendency' to use jh to express Iranian z, e.g.
vajheska, erjhuna, marjhaka. But it is only found once to express the
voiced internal -s- in majh\€'\. In Niy'a forms like divajha, dajha,
beside divasa, ddsa, are commoner.
Internal -j- appears as -y- just as in Niya. Also internal -c ayaria .
sahayara. The latter in Niya apparently becomes k, f (representing
-Z-). praknra < pracura, vajida < vdcita “read”. The Kharosthi
Dh.P. always has -y- here — koino (i.e. koyino) for kocinah, so that the
k (j) is apparently' a peculiarity of the Niya Prakrit.
In inflection the evidence is much less e.xtensive than for phonology,
owing to the jejune nature of the texts. It appears, however, that
the process of decay and innovation has not been pushed so far as in
the Niya Prakrit. For instance, the original forms of the nominative
-e and -o are preserved, whereas in Niya, though there is evidence to
show (see above) that the nom. sing, originally' ended in -c, it is now
confused with the accu.sative, both ending in -a. Sten Konow {Corp.
Inscr. Ind., vol. ii, p. cxii), after examining the places where -o occurs
and those where -e occurs, comes to the conclusion that -e prevailed
in the regions west of the Indus, while -o is usually to be found in the
THE DIALECTICAL POSITION OP THE NIYA PRAKRIT
425
districts to the east of it. The original home of the Niya Prakrit
would therefore on this ground have to be sought in the area west of
the Indus.
The locative sing, in -ami agrees with the Mya dialect as opposed
to Asokan -aspi, -asi.
A syntactical point found both in these inscriptions and in the
Niya documents is the habit of stringing large numbers of nouus
together in quasi-compounds.
Examples are (from Konow, op. cit., p. cxv) mahadanapati
Patikasa, erjhuna Kapasa, maharaja rajatiraja Hoveskasa ; (from
Niya) 133 priya nivasaga Svaneyasa “ to his dear neighbour Sv'aneya ”
575 tnaya rajadivira srmimmna Dhamapriyena By me the royal
scribe, the monk Dharmapriya
Certain technical words are common to both sets of texts, notably
the terms used in giving dates, ksuna “ time ” and saste “ day ”.
On the whole, then, it is quite clear that the correspondence is
very close, much more than with the Asokan texts. There are very
few points, indeed, where they seriously diverge. The change sv > sv
is peculiar to Niya, but might easily have developed out of sv at a
quite late date. In inflection the Niya dialect shows quite a number
of innovations not found in the other Kharosthi Inscriptions. But
these, too, indicate merely a more advanced stage of linguistic develop-
ment rather than actual difference of dialect. Such are the confusion
of nominative and accusative mentioned above and the development
familiar in modern Indo-Aryan of a transitive preterite from the past
participle passive, didemi “ I gave ”, drithesi “ you saw ”, etc.
Dividing the dialect of the Indian inscriptions according to the
nom. sing, in -e or -o, we may class the Niya language as being very
closely connected with the -e dialect.
Ill
The only hterary text in the North-Western Prakrit is the
Kharosthi Manuscript of the Dhammapada. This text shows strong
influences of the original dialect from which it was deriv-ed, and before
evaluating it for dialectical purposes this has to be discounted.
Examples of this element are : —
The treatment of ks as kh : bhikhii, khano, cakhmna, pratimukhe,
punapekha. The regular dialectical treatment is ch — chaya, bhichavi,
vichitani. No distinction is expressed, as in the Niya documents,
between ch = ks and ch =ch. Of interest is aveha 31 which
v'OL. vin. PASTS 2 AXD 3 . 28
426
T. BURROW
= apeksa. The change ks > h (after a long vowel) is characteristic
of Ardha Magadhi. Obviously the original text from which this was
adapted cannot have been Pali, but, if not Ardha Magadhi, at any rate
something connected with or influenced by it. This points to the
existence of versions of the Buddhist scriptures in Prakrit dialects
of which no direct trace remains.
Further assimilation of r is often to be ascribed to the same source :
silavata B 24, udaga B 30, katta (=karna-) B 34, bhayadasima B 32.
ahha (= abhrdt) A^ 3, dhama A® 4, as opposed to the regular dialect
forms niarga B 6, darsana C"'° 37, matrena B 24, etc.
vasa appears with assimilation of -rs- in C™ 18. 19, as commonly
in the inscriptions of the Xorth-M est. On the other hand, we find
varsa C™ 2, as in Niya. In some cases it is not easy to say whether we
are dealing with a form borrowed from the original text or the genuine
dialect form, namely when both forms are found elsewhere in different
XorthAVestern documents. Thus vasa, which occurs twice (see above],
we might have thought to be the genuine dialectical form because
that appears both in Asoka and the later Kharosthi inscriptions.
But at C™ 2 we find varsa as in the Xiya documents. Similarly we
find apa for alpa. Ip is assimilated as pointed out above in the North-
IVestern versions of Asoka but preserv'ed in Niya. Since we cannot
be sure that the assimilation in apa of the Kharosthi Dhammapada
is not due to the original text, we cannot be quite secure in takmg it
as evidence that this text represents a dialect in which that
assimilation had taken place.
Other examples that may be mentioned are : suyi = suci. probably
because the adjective suci- existed no longer in the vernacular of the
North AYest ; anasava (= andsrava-), contrasted with the regular
treatment of sr in anavasutacitasa. Further may be mentioned
numerous inflexional forms which were current when these verses
were composed, i.e. in the early Buddhist period, but later were not
used in middle Indian, for instance the ablative in -a {= at) abJia
(abkrdt), gen. sing, of stems in -avt in -ato (by this time everywhere
-antasa), futures like tivehisi C'° 22, aorists etc., most of which had to
be preserved, of course, for the sake of the metre.
Subtracting the.se elements the language displays considerable
similarity to the dialect of Niya. The weakening of internal consonants
has proceeded on much the same lines, although the alphabet expresses
it much less satisfactorily. There is, for instance, no difference between
s and s. cli and ch. and this may easily be due to defects of the
THE DIALECTICAL POSITION OF THE NIYA PRAKRIT
427
alphabet. Internal -g- appears WTitten as -h-, e.g. urako, kuiakare,
etc. This is curious, but there is no doubt that it represents what in
the Niva documents is expressed as g, because just as Niya g it tends
to be confused with y. Thus udaka is written for udaya (B 13, C''“ 18)
and dhoreka for dhoreya 0''“ 37. Similarly in Niya apramego is WTitten
for aprameya and I'yaga for vyaya. The omission of internal consonants,
which in Niya occurs only in suffixes -ika, -aka. -uka (e.g. agamduva).
is somewhat further developed in the Dh.P. Thus, besides cases of
omission in suffixes, daruva = daruka C'" 31, athagio 4, we have
forms like roa = roga, hhoha = bhoga, with -h- inserted in the hiatus.
Similarly -t- is omitted in abhai B 7 = abhati. pha.fai A® 10 = sprsati,
although it is usually preserved ; sali is wTitten for sayam in 0'“ 12,
showing that -t- was coming to be very lightly, if at all, pronounced.
The appearances suggest that the Dh.P. represents a slightly more
advanced stage of development phonetically than do the documents
from Niya, and the sporadic omissions of -g-, -t-, etc., probably mean
that in the spoken language of the people connected with the
manuscript the tendency had become general, but the traditional
orthography only allows it to appear occasionally in the text. From
this we would conclude that the manuscript was later than the Niya
documents.
Further common characteristics are the changes ny> >1, sr > sp,
and sr > s ; ms becomes mts, but the anusvara is consistently not
written in this manuscript, so it appears as t$, satsara A^ 6. ahitsai
A^ 8. It was wrongly transliterated fis by Senart.
A marked characteristic of this text is a tendency to represent
V by w, e.g. bhamanai < bhavandya. nanta < ndvatn, and even pu.saviva
ior pmpam iva. In the Niya Prakrit this occurs regularly in one word —
games “ to seek ” ; also amecJiita in the verses quoted from the
Dhamma P., No. 510. Perhaps cimara 149 may be < clvara.
A tendency common both to the Niya Prakrit and the Kharosfhi
Dh.P. is the voicing of unvoiced stops when preceded by a nasal, and
the omission of voiced stops in a similar position. This was treated by
J. Bloch in the article mentioned above, showing that it occurs in
modern times all over the North-West. Examples are (a) paja 5
sagapa < satikalpa, etc.; (6) htnaru “elephant”, iidiimaru, etc.
In Niya we find (a) upasamghidavo = upasank-, samghaUdavo from
samkalayati “ to collect " , gamdavo (so written fourteen times as against
gamtavo twice). No examples are found in the case nc. pamca, etc.,
being invariably found. We cannot be sure how this is to be accounted
428
T. BURKOW
for because there is a constant tendency due to the phonetic character
of the native language to unvoice voiced stops in Indian words which
might possibly mask a change like this. Still the fact that -mj- never
occurs would speak for the change not having taken place in this
dialect.
(6) gamnavara “ treasurer ’’ (Iranian *gan]a^ara-), chimnati “ he
cuts ” < *chindati, which has replaced Skt. chinatti, bJiimnati,
bamnanae “ to bind But forms like batndhitaga 660 are not
uncommon. The development has not taken place in Niya in the case
of nd (as it has in the Dh.P. panito < pandita, etc.). We always
find pini^, damda, etc. In the case of ->»6- the only example available
is hastama “ quarrel ”, which is an Iranian word from the base starnb-
(cf. BSOS., vii, p. 788).
On the whole, then, the Kharosthi Dhamma P. is much the more
consistent in carrying out this change and it may be taken as an
indication of dialectical difference.
There is a tendency, sporadic but quite common, to change e
into I, e.g. viranesu, pramodia, uvito, sarvi. This change is not found
apparently outside Khotan (see below). The question, therefore, which
naturally arises, whether the manuscript was copied in Khotan or
imported from India is settled by this point. It must have been copied
in Khotan, perhaps more than once, to account for this specifically
Khotanese change.
There is one point in which the Dh.P. differs from the Niya dialect
and the later Kharosthi inscriptions, but agrees with Asoka, and that
is a tendency (not universal, but common) to transpose r. Examples
are drugati, drumedhino, driigha, pravala.
The specific Asokan change sv (and sm) > sp is not found here,
as, indeed, in any other text outside Asoka.
In inflection the most important difference is the treatment of
the nominative singular. The Kharosthi Dhammapada represents an
-o dialect; Niya, as shown above, an -e dialect. Judging from the
division of these forms geographically, which is mentioned above,
the home of the Dhamma P. dialect is to be sought to the east of the
Indus.
As we have seen at Niya, no distinction is preserved between
nominative and accusative. This would seem to have been the case
at the time the present copy of the Kharosthi Dh.P. was made.
Because, there reigns complete confusion, as seen from the following
hst of examples : —
THE DIALECTICAL POSITION OP THE NIYA PRAKRIT
429
N. S. in -0
„ -u
„ -a
Acc. S. in -u
,, -u and a
N. Acc. N. -o
„ -n .
j, -Cl .
magana athagio setho
nai kalu jyramadasa
asoka sainojana
so itu loku ohaseti
dhamu sucarita cari
bhayo C'“ 7
pranoti paraniii siikhu; abhaya namu sa disa
apratnadu mmtapada, pramado niucuno pada
This state of affairs seems most naturally to be explained by
assuming that the people among whom the document was current no
longer distinguished the two cases. This is in agreement with the
conclusion above, that phonetically there were indications of a more
advanced stage in the popular speech.
Another important distinguishing mark is the locative singular.
That, as we have seen, is -aspi or -asi in Asoka and -ami in the later
Biharosthi inscriptions and Niya. In the Kharosthi Dh.P. we
frequently find a form in -asa, e.g. uthanaalasa (cf. A* 6, A® 6, A® 13,
B 30, 38, O'"® 3, 28, 30). The form as it stands is identical with the
genitive singular but the syntax in all these passages demands a
locative. It must therefore represent an original -asi. How the final
-a instead of -i is to be accounted for is not easy to say. We might
think of a phonetic process resulting in the confusion of final vowels,
but there is no further evidence to support this. The only other
explanation is that it was due to ignorance of a form -asi among the
people among whom the text was current. Not knowing the form
they would easily confuse it with the genitive in -asa. But how did
this occur ? We might assume that original text (in which already one
Ardhamagadhi characteristic has been pointed out above) had locatives
in -a{m)si, and that these were not current in the dialect into which
it was translated. On the other hand, locatives in -asi are attested
in the North-West in Asoka, though not in the later Kharosthi
inscriptions and in Niya. Since, however, the Dh.P. shows quite a
number of differences from these latter texts, it would not be at all
surprising to find the locative in -asi in it. The difficulty in that case
is to account for the confusion with the genitive which there seems
no reason to explain from phonetic causes.
A further point of comparison is afforded by the indeclinable
participle in -ti, which as in Niya and Asoka is in -ti : upajiti 0"'° 44,
pramayiti A® 3, parivajeti A® 8.
430
T. BURROW
To sum up, the main differences between the two dialects are not
many but striking where they occur. ATe notice in the Dh.P. a more
pronounced tendency to voice voiceless stops after nasals and omit
voiced stops in similar positions, a more pronounced tendency to
change -r- into -/«- ; the transposition of r ; the nom. sing, in -o,
and possibly the locative singular in -asi. In all these the group of
later Kharosthi inscriptions having nom. sing, in -e agrees with Niya
rather than with the Dh.P. In one point, the transposition of r, the
Dh.P. recalls Asoka, possibly also in the locative -asi. The evidence
leads us to seek the original home of the Dh.P. east of the Indus.
Its date is, if anything, later than the Niya documents.
IV. The Dialect of Khotax
One document out of the collection — 661 — is written in a different
script and dialect from the rest. It was found at Endere, but since it
is dated in the tenth year of Anjida Siinha, King of Khotan, we may
take it as representing the administrative language of Khotan, just
as the Niya documents represent the administrative language of the
Shan-Shan kingdom. The date of the document is not settled, even
approximately.
Since the language looks noticeably different from that of Niya,
it will be profitable to analyse the differences in detail.
e and o become T and u. Kali sagaji { — sakcise), uvadayi ( < uvadaye
< upaddya), the optatives vidiyadi (Niya veteyati), uthaviyadi, cudiyadi,
dhinadi “ he gives ” (Niya denati).
o > M in cudiyadi (Niya codeyati). The same characteristic appears
in Khotanese Saka (S. Konow, Snla Studies, pp. 19, 20). It appears
in this document in the Iranian title Hinajha = aTpaTyyos
{BSOS., vii, 514). Its occurrence in the Dh.P. was mentioned above
as indicating Khotane.se influence in that document.
In the instrumental the -c is treated differently, becoming d, e.g.
mulydna, tdua. The same appears in the 3rd sing, of the causative,
madrddi = mamtreti. It is not obvious whv there is this difference.
We might think of it as due to their being terminations and
therefore less .strongly accented, resulting in a shortening of the
e to e. But the -e of the locative singular is treated like the ordinary
e — lioli, .sayaji.
Initial d- is regularly aspirated, dh-, dhivajha “ day ”, dhilu
“■ gave ”, diiadu = damdam. It must be due to native Khotanese
pronunciation.
THE DIALECTICAL POSITION OF THE XIYA PRAKRIT
431
Internally the change does not occur in the only example available,
vivadu = vivddam.
Between vowels -t- always appears as -d-, harctdi, hidd. grahidu.
uvagadu, dhinadi, syadi, etc. This being the case, we can conclude
that since the past participle of dd- appears as dhitu we are dealing with
a double consonant, i.e. *ditta-. Niya dita therefore is not directly
derived from Aryan *dita-, i.e. data, but another of the numerous
new formations that appear in India. The form *ditta- for the North-
Western Prakrit is also shown by TorwaU dit " gave ". A single -i-
would not have been preserved, cf. Torw. gd " went
The tendency to voice internal -I- was, of course, general in the
North-West as everywhere. In the Niya documents, however, the
development is masked on account of a general tendency to confuse
voiced and unvoiced sounds. The native language of Shan-Shan, as
I have shown in JRAS., 1935, pp. 667 ff., was like Tocharian in being
devoid of voiced stops. Consequently they confuse them in Indian
words. The fact, therefore, that a similar confusion does not exist
in this inscription is of great interest, because it demonstrates that the
native language of Khotan (pre-Saka) was quite different from the
probably Tocharian language of Shan-Shan. Naturally one document
does not enable us to say what it was, but it is satisfactory to be able
to fix a boundary to the extension of Tocharian at this time.
The group st which otherwise is assimilated in all varieties of the
North-Western Prakrit is preserved in asti “ 8 The final vowel is
also surprising, recalling Saka hastd (besides Jiasta). On the other hand,
we find the usual assimilation of -sir- in Utah “ camel ”.
The group -sy- is always preserved in the genitive singular
°simhasya, etc. Of course, this is merely a case of imitating Sanskrit.
In inflection the most important difference is that in this document
the distinction between nom. and acc. is stiU preserved, whereas in
Niya it is quite obliterated. The nominative appears either as -d or
-ah, the acc. as -o or -u.
Examples of the nom. sing, are ; asti nmnusd mgaragd " There is
a man of the city ", asti umyi Utah tanuvagah, so Utah Vagiti Vadhagasya
tanuvagah samvritah “ That camel has become the property of V. Y.”
Examples of the acc. sing, are : so uto vikriuami " I am selling that
camel ” ; Vagiti Vadhaga hiravasiso mulyo mam dhitu " V. V. gave
the whole price (the construction of masd in this sentence is not clear) ;
vivadu uthaviyadi “ should stir up a dispute ; dhadu dhinadi “ pays
a penalty ”.
432
T. BUBROW —
This being established, the phrase so Utah aphihanu haradi is to
be translated “ That camel carries apKinanu ” and not (as wrongly
taken by P. S. Noble, BSOS., vi, 451, and by me, BSOS., vii, 515)
“ Aph'inanu carries away that camel
Of these two forms of the nom. sing, -ah if it really represents the
visarga must be simply a Sanskritism. The -a is some kind of e-sound,
as shown by the forms tana and mamdradi. It is to be classed therefore
with the e-dialects along with Niya. WTiat is surprising is that we
seem to have here a treatment of the -ah of the nominative sing, distinct
from all other Prakrits. Everywhere where it becomes -e the -e is
not distinguished from the ordinary -e, of the locative singular for
instance. Here we have i < e in the locative and d < ah in the nom.
sing. It is a pity there is not more information, e.g. in the form of
ablative singulars (Niya -ode) on the point. The only other form
containing an original -as is puradu, with development in the o-
direction. This, however, is of no significance, because regularly in
those dialects which have -e in the nom. sing, -o occurs in adverbial
forms like these (J. Bloch, BSOS., vi, pp. 291 fif. ; L’Indo-Aryen, p. 8).
The change -am > -u in the accusative singular is interesting.
In Niya it appears as -a as does the nominative, but there are traces,
in the pronominal forms, of the change -am > -u, namely ahu “ I ”
amahu tumdhu (< asmabhyam, etc.) tuo “ you ” < tuvdm.
In yatha rajadhamu syadi we apparently have the neuter singular.
The document fimther throws considerable light on the history of
the new preterite which the Niya dialect has formed out of the past
participle passive. They say, for instance, ditemi, ditesi, dita, ditama,
ditetha, ditamti for “ I, you, etc., gave ”. It is not clear from the
Niya language itself what the -a in dita (3rd sing.) represents. The
Khotanese document, however, shows that it was originally the neuter
singular that was used here. We find, for instance, tasya utasya
kida Vagiti Vadhagd niravasiso mulyo masd dhitu Khvarnarsasya
grahidu “For that camel V. V. gave the whole price and it was received
by Khvarnarse ”. Here we find both the original construction
{Khvarnarsasya grahidu) and the new one {dhitu with the nom.
sing.) side by side.
Development in a similar direction is shown by the Niya documents.
We get : —
(1) Constructions with the subject in the instrumental 16 maya
maharayena Feta Avang camkura Arjunasa picavida “ By me the great
king P.A. w^s hand^ over to the carnkura Arjuna ”.
THE DIALECTICAL POSITION OF THE NIYA PEAKRIT
433
(2) The nom. sing, supplants the instrumental in conjunction with
the past participle edasa rnahuli stri Raniasri unidi gida “ His wife
took a woman Ramasri as an adopted child
(3) The instrumental is used as a nominative even with the present
tense. 622 muharayapiitra kola Pumnabalena lihati “ The king’s son
hala Pumnabala writes ”.
Whether the application of the personal endings to the past
participle had taken place in the Khotan dialect as it has in Niya it
is not possible to say. No example happens to occur in the
present text.
The intransitive samvritah agrees in gender and number with its
subject. In Niya transitive and intransitive verbs are treated alike.
The differences between the two varieties of Prakrit may therefore
be classified under the following headings : (1) Special developments
in Khotan, e>i, o> u, d-> dh-.
(2) Special developments in Shan Shan, notably a widespread
tendency of unvoicing and confusion of surd and sonant.
(3) Differences due to the fact that Prakrit of Khotan is more
archaic than the Niya Prakrit. They are the preservation of the
distinction between nom and acc., and a somewhat less advanced
stage in the development of the new preterite from the past participle
passive.
(4) There remains the possibility of original dialect differences
in their Indian sources. One document naturally does not provide
much to found hypotheses on. There is asti “ 8 ” where there seems
to be preservation of -st-. This is certainly not what we woidd expect
because otherwise that group is assimilated in all varieties of the
North-Western Prakrit. The final -i also is difficult to account for.
The possibility must certainly be taken into account that a form
like this is due to the influence of Khotani Saka, which has hasti
for “ eight ”. When the Sakas came to Khotan is not known, but there
is clear Iranian influence in this document. The king has an Iranian
title hinajha {*lilndza-). The man whom the document is about has
a name which is undeniably Iranian, Khrarnarse (= *xvar-narseh,
BSOS., vii, 789). So that it is natural to assume that Iranians,
presumably Sakas, were already here. That being so, Iranian influence
seems more probable than actual preservation of -st- in the Prakrit.
Also we get the usual assimilation in Utah.
If we are going to consider the possibility of Saka influence, we
might also see it in the nom. sing. An d is otherwise unknown in Middle
434
T. BURROW —
Indian, but we do get a similar modified a in Saka in the nom. sing.
gyasta, etn. Since there is no trace of such a form inside India, from
which the language must have been imported, and since everywhere
the -e < ah of the noni. sing, is not distinguished from original e, there
is hardly any possibility left except that it was invented in Khotan.
Similarly tdyia = Saka ttdna, madrddi = Sak. dydhata, etc. The
accusative, too, in Saka is -u, gyastu, etc., just as here.
If we assume that these things are due to the influence of Saka,
then there remains no difference between the two dialects which
need to be attributed to their Indian source. Basically they are the
same language and the fact that they look so different is due almost
entirely to recent developments in their Central Asian envdronment. The
Niya dialect, as we have seen, coincides pretty closely to that part of
the Kharosthi inscriptions of N.W. India, which have the nom. in -e.
These for the main part are concentrated in that area to the west of
the Indus round Peshawar, which was the centre of the Kushan
dominions. It is this official language, then (as one would have
expected for historical reasons), which has been transplanted to Khotan
and Shan-Shan, taking on a slightly different form in each.
ilost of the phonetic peculiarities of this dialect reappear in the
modern Dardic languages. A few of the phonetic developments are
particular to Torwali, namely : —
(1) sv > SI', Niya svasn, Torwali su. In the closely related Garwi
ispo, as once in Niya, priyaspasuae. On the other hand, Sina has
sd with the s preserved.
(2) st> > sp. Niya aspa, speta, Torwali pais < {s)pasu < h'asrii.
This change is also found in Sina, dspo but not apparently in Kasmiri,
chygt'^ ■■ white ".
(3) sm > m, Niya amahu, and locatives in -ammi, Torwali mo
'■ we ” < (a)maku. On the other hand both Sina and Kasmiri show
a change to s (as in Asoka to some extent and the Kharosthi Dh.P.),
KA as' '■ we ", S. nsri ‘‘ ower ".
These are the mo.st important points establishing close relationship.
Other forms of interest are : Torw. biz, Niya bhija “ seed " (with
unexplained bh-). Llsually in the North-Western Prakrit -j- became
-y-, maharaya. The in this form is probably to be e.xplained by
gemination. b7ja > blya > biyya >, bijja > bhija > Torw. biz.
Similar is the case of Niya ija here ”. Skt. iha > Asokan ia > iya
> yyya (with additional emphasis natural in such a word) > ijja
> ija.
THE DIALECTICAL POSITION OF THE NIYA PRAKRIT
4;55
In Niya Skt. dv is treated differently according as to whether it
was originally dv or duv-. Thus dvara but biti. Similarly Torwali
has bi ” second ”, der ‘‘ door ” (though the latter looks as if it may be
Persian). On the other hand, they disagree in the word for " 12 ’’ ;
Niya badasa, Torw. dvdL
One point which appears in Torwali but not apparently in Niya
(as mentioned above) is the transposition of r. Torw. pk ” high ”
< *drigha < dirgha-, pibal “ thin ” > *dnibah, > durbala.
Further there is no trace in any of the Dardic languages of the -t-
which develops between ms in mdintsa, etc. Torw. )nds, Sina mos.
It might be going too far to say that Torwali is the direct lineal
descendant of the Niya Prakrit, but there is no doubt that out of all
the modern languages it shows the closest resemblance to it. A glance
at the map in the Linguistic Survey of India shows that the area at
present covered by “ Kohistani " is the neare.st to that area round
Peshawar, where, as stated above, there is most reason to believe was
the original home of the Niya Prakrit. That conclusion, which was
reached for other reasons, is thus confirmed by the distribution of the
modern dialects.
On the Alphabetic Notation of Certain Phonetic
Features of Malayalam
By Raymond T. Butlin
To search, for “ unity ” and system ” at the expense of truth is
not, I take it, the proper business of philosophy, howe-cer universally it
may have been the practice of philosophers. — G. E. Moore.
I T is proposed here to offer some considerations on the representa-
tion of certain Malayalam ^ sounds. It is not the purpose
of these notes to present a detailed laboratory account of the acoustic
and physiological phenomena to be symboUzed, but while characterizing
these phenomena sufficiently for the reader to appreciate their general
nature, rather to consider certain specific phonetic circumstances of
their occurrence and the significance of these in the elaboration of a
notational system.
[A brief indication is here given of the sound-values which
the reader should attach to the symbols hereafter employed, in
order to enable him to make some verbal response to the visual forms,
and in the case of those familiar with the language, to assist in the
identification of the words.
Yoivels and Diphthongs. — The precise value to be given to the vowels
and diphthongs is irrelevant to the present discussion. Twelve symbols
will be used, viz. a, aa, e, ee, i, ii, o, oo, u, uu, a, ai. (A symbol is not a
letter : aa and a are two distinct symbols.) It wiU be found sufficient
for the present purpose to read these symbols in accordance with inter-
national phonetic usage, length being represented by double letters.
Consonants. — The reader will find it sufficient to attach to the
following symbols the values described in J. R. Firth’s “ Short Outline
of Tamil Pronunciation ” ^ ; p, pp, t, tt, J, c, cc, j, k, kk, m, mm,
!]_, i|_i|_, p, pp, q, qq, 1, 11, |_j_, v, vv, y, 3 ry. Those who are unfamiliar
with this work may, without prejudice to the main considerations of this
article, interpret these symbols, together with b, ^h, g, jj, s, q, J, and h,
in accordance with international phonetic usage, subject to the following
modifications : c, cc, j, jj are palato-alveolar affricates, g is used for the
^ Our informants were Travancore Brahmins.
^ Appendix to Arden's Grammar of Common Tamil (new and revised edition),
published by the Christian Literature Society of India, 1934.
438
E. T. BUTLIN
International Phonetic Association / is retroflex, y has its English value
in i/es.
The following symbols require special explanation : tt, n, nn, n, nn,
r, r, i, kk.
tt. The primary articulation of the sound represented is alveolar and
tense, with slight secondary palatalization.
n, nn. To be read respectively as short and long nasals with primary
dental articulation and slight secondary velarization.
n, nn. Short and long nasals respectively with primary alveolar
articulation and slight secondary palatalization.
qq. Pre-velar articulation, intermediate between Ji and q.
r, A short alveolar trill Tsith slight secondary velarization.
r. A denti-alveolar or dental trill with secondary palatalization.
The symbol should be given the value assigned by Firth (v'.s.) to J.
A retroflex frictionless continuant.
Should be interpreted as a tense voiceless palatal plosive.]
Our observations are based upon an examination, on the phonetic
level, of the circumstances in which certain soimds occur in a number
of selected Malayalam words. These sounds will, as a result of our
analysis, be symbolized by (t, tt, tt, |;^), (m, mm, n, nn, n, nn, p,
pp, qq, q, qq), (1, 11, ],]_), (r, r, p), (k, kk, ^). Both from the point
of view of orderhness of exposition and of typographical con-
venience we shall find it advantageous to treat the material in
groups, as indicated above by the use of brackets. It should be
stated that this methodological device has been so chosen as in no
way to affect our conclusions.
We shall examine in detail the alternances occurring within these
groups in initial, intervocalic, medial, and final positions. The material
will not, however, be presented in this schematic order, but in
diminishing sec^uence from the series presenting the maximum, to
that presenting the minimum number of alternative terms. ^
In certain cases it has been possible to establish a complete series
on the basis of a single term alternance in complete words. Such
series are prefixed with an asterisk.
Important as such cases are, it w'ould, however, be unpossible to
devise an adequate notational .system exclusively on the basis of
entire words, since, of the theoretically possible number of single-
term alternances in complete words, only comparatively few are
actual. No doubt further investigation will reveal such series which
the writer has not yet had the good fortune to discover. Neverthe-
* The exprebbion “ alternative phonetic terms” is used by J. I!. Firth in The Use
and Distribution of Certain English Sounds (English Studies, in.3.5).
NOTATION OF MALAYALAM
439
less, it is certain that no amount of research would reveal single-
term alternances in all the contexts chosen for the present purpose.
In the majority of cases, therefore, it has been necessary to employ
words exhibiting multiple alternance, by extracting from them
appropriate isolates in the form of particular phonetic sequences
exhibiting single-term alternance. The specificity of the isolates
selected for the present purpose is such as to enable us to devise
a notation making the fullest use, compatible with unambiguity
and practical convenience, of contextual conventions involving
contiguous terms. By increasing the degree of specificity, such
contextual conventions could be considerably reduced.
TABLE A
Series I. Intervocalic
1
t i
tt
1
tt
1
i t
u
*pati
husband
*patU
hood of snake
♦patti
glued
»PBti
a step ^
*patp
dog
•kuti
a jump
*kutti
stabbed
*kutti
tent-peg
•kup
a drink
•kntti
boy
koti
greed
kotti
Stung
kop
flag
kolp
struck
paata
path
paatta
cockroach
paafa
scum '
1 kota
1
umbrella
kopa
1
basket
Series II. Medially, preceded by Homorganic Nasal
t
tt
t
panta
niintuka
ball
to swim
tantte
makantte
entte
one's
son's
niy
kai]_^u
niintu
saw
became long
Series III. Medial Groups (other than those included in
Series II and IV)
t
t
vaastaTam
truth
kaj|;am
troublesome
paqti
column
JaJtipuurtti
60th birthday
' The term has been adopted from Professor H. Levy.
440
R. T. BUTLIN —
Series IV. IHedially, with r or y
tt
raattri night
varttamaanam news
sattyam truth
marttyan man
Series V. Initial (including Initial Groups)
tanta
tekka
tiircca
tu^al
tyaagam
tyajikkuka
stanam
father
south
decision
a chain
a sacrifice
to discard
breast
It will be seen that in Table A the series exhibiting the maximum
alternance occurs intervocalically and contains five terms, while
medially with preceding homorganic nasal we have a three-term series,
in other medial groups a series of two terms, and a single term
initially and in medial groups with r or y.
It is evident that the number of symbols necessary and adequate
for the representation of a series of alternative terms is in direct
proportion to the number of such terms. The maximum number of
symbols will be required in the series exhibiting the maximum
alternance, the minimum in the minimum series.
Now it is important to realize that no useful purpose whatsoever
can be served by seeking a chimerical unity between a term in a major
series and one in a minor series, or even between the terms of two
different series containing an equal number of terms, since to do so
is to leave out of account the relevant context of their occurrence and
so to invalidate any conclusion.
Strictly speaking each term might be distinctively symbolized,
but for the practical purposes of notation it is desirable to practise
some measure of symbol economy. This may best be effected if the
choice of symbols in a minor series be determined by an empirical
comparison of the sounds in question to similar sounds in a major
senes (subject, however, to a provision to be explained subsequently).
Series I. Intervocalic
NOTA'I'ION OF MALAYALAM
441
B*
S>
o c -o
BO s U
S O
<9 C ^
s “ g
O "
«•
s-l.
B> c*-**
2 ® S
B O 0
S 8 >
B*
11.1
E S
0
1
B «• C
B< B*
g* 5* a
S“ g.
SM-
e.
B.
o 5 ^ E
«2 o
S = i s
- c .s
« V
o ^
.a &
• g* c,
c, B, e. e.
e. c. 2 B
B 9 <t
B.J9 M e%
*= ^ s ^
•T3 ^ .3 «
c e£
S
B*
r* S'-
& B* B* =;
B* B* B* 3
s B <• 2
o 8
*g ^ g
S e s Q.
- f-® S
B*
.a
- S 0.8
- Ss 2*
a A s fik
a-
.B e
= ^*2
S 2 «
ill
.M a B.
e
O a O ■£
* «-g ^
B <3
8
B
B
§ 2 i
B B ? •
B a O B
B o- 5 3
M B 6 ^
B
B
T* M ti "3
E I i ^
= “ i
1 »
-» I
B e a 8
B 0 B -8
W Ck « B
B
B
s s s
g
S|
ill
I S 2
B
&i e ^
S .E 'b
■S s
4J
8
• 8
•-3 1
ill
VOL. vm. PARTS 2 AND 3.
29
442
R. T. BUTLIN —
Thus tt is used in medial consonant groups in the foregoing examples
on the ground that empirical observation shows the sound to be
almost identical with that sjunbolized by tt in the intervocalic series.
Absence of approximate acoustic identity does not necessarily
preclude the use of identical symbols for two terms in different series.
Thus, although there is a difference of tensity and voicing between the
terms of the initial and intervocalic series symbolized by t, no
ambiguity can arise if the appropriate convention be adopted. It
woxild, nevertheless, have been equally legitimate to employ a
distinctive symbol, say 8, for the term in the intervocalic series ; and
such a proceeding might, for certain pedagogical purposes, be
advantageous. Since, however, Malayalam is a written language, and
that in the orthography both terms are represented by identical
symbols, for general purposes it is perhaps more convenient, without
implying any relationship, to use similar symbols in both cases.
Series III. Other Medial Groxips (excluding those in Series IV)
m II q
samsaari^uka to talk ■ anyan stranger , paqti column
tinmaan for eating
Series IV. Initial
m
1 n
maattnka to move
1 naaraqija
lemon j jiaan
I
maram
tree
1 ninnu
stood nannal
we
bamboo
1 niinti
!
swam j jioji moment
1
Series V. Medially, after r or
nun nn
nirmmikkuka to appoint ' tiimnu finished
^ taa][ima lower
NOTATION OF MALAYALAM
443
Series VI. Final
m
n
mai|_ani
smell
pasryan
boy
ikkaalam
this time
veejan
hunter
addeebam
he
cemmaan
cobbler
Series VII. Initial Group
n
nyaayam justice
The nasal group presents a maximum series of ten terms inter-
vocalically. The series with following homorganic stop presents six
terms. It will be observed that for the second, fifth, and sixth terms
of this series, symbols have been employed which do not occur in the
notation of the maximum series, viz. n, ji, q.
Now, although the sounds thus symbolized do not occur inter-
vocalically, it would be possible, by establishing a contextual
convention as to length, to employ either nn, jiji, q;), or n, ji, q in
both series. Such simplification nevertheless appears unnecessarily
TABLE C
Series I. Intervocalic
1 U I ll
palaka plank pallakka a litter kaj^a weed vel_l_am water
baalika girl alii bud ka^a|_i kind of banana ; pa[[i church
mula breast mulla kind of plant mu]_a bamboo i mu]^|a thorn
Series II. Final
1 I
pakal daytime avaj^ she
mayil peacock kavi\ cheek
vaQaal perhaps aa^uka]_ persons
444
B. T. BUTLIN —
Series III. Initial
1
laatam horseshoe
ookam world
lejja shyness
Series IV. Medial Groups
1
taalpparyam meaning
kalppam kind of tree
schematic and is of no practical convenience. It is gratuitous to assume
that in all cases contextual conventions are necessarily preferable
to additional symbols.
Similar arguments have been advanced for the use of either h or
q for both h initially and q finally in English. It is interesting to note
that by applying the method here described, such fruitless discussions
are avoided. The intervocalic series in English contains both h and
q, while the initial (a minor series) contains h but not q, and the
final (also a minor series) contains q but not h. Eeference to the
intervocalic (major) series will suggest the use of h for a similar
soimd in the initial series, while reference to the same series will suggest
the use of q for a similar sound in the final series.
Tables C and D are presented without comment, since they involve
no new problem.
TABLE D
Series I. Intervocalic
■ 7
r
r
■
1
*kara
sap
*kara
coast
*ka];a
pimt-pole
kiiri
tore
kiiri
mongoose
ki];i
bundle
coora cooked rice
coora
blood
koo|;i
fowl
mara
screen
pura thatched hut
pu^a
ma|^
a stream
rain
NOTATION OF MALAYALAM
446
Series II. Medial Groups
r
I
iircca uneasiness
yaattra journey
tarkkam dispute
suuryan sun '< covvaa];cca Tuesday
kaaryam fact |
viiryam bravery taapma lower
Series III. Initial
r
r
raantal
lantern
raattri night
raai],!
queen
ruci sense of taste
ravukka
cotton or silk
jacket (women’s)
roomam hair
Series IV. Final
r
payar
peas
kayar rope
malabaar Malabar
Series V. Initial Groups
r
kramam gradual
^ramiccn tried
There remain the velar and palatal unaspirated plosive alternances
(Table E). It should be noticed that previous writers have failed to
observe the alternance value of the third term of the intervocalic
series, Aiyyar,^ for instance, treating such cases as variants of kk “in
association with palatal vowels ”. The following examples leave no
doubt that ^ is a discrete term.
* L. Vishwanatha Ramaswami Aiyyar, A Brief Account of Malayalam Phonetics.
446
R. T. BUTLIN
TABLE E
Series I. Intervocalic
k
kk
akam inside
pooka go (imper.)
maka]^ daughter
akkam
pookka
makka|_
kaakkuka
arakka
arakkuka
a numeral '
gait'
children
to guard
sealing wax
to saw
avan vikkaan he is going
pookunnu to stammer
kaa^uka to ripen
ara^a grind (imper.)
arakkuka to disguise
avan vil^aan he is going
pookunnu to sell
Series II. Initial (including Initial Groups)
k
karam
kariijijaali
kiiri
tax
name of tree
mongoose
kramam
gradual
Series III. Medially, preceded by Homorganic Nasal
k
maqka virgin
vagkan fool
paqkajam lotus
Series IV. Other Medial Groups
kk
tarkkam dispute
markka^am monkey
karkka^akam name of a month
Conclusion
Avoiding that crude hypostatization of the visual symbol which
vitiates so much contemporary research, it has been possible, by
a methodical examination of certain sounds in their relevant contexts,
to establish a systematic, unambiguous and manageable notation.
NOTATION OF MALAYALAM
447
Although not an end in itself, a notation is an indispensable
instrument both in broader linguistic studies, whether descriptive or
historical, and in pedagogical practice.
In itself a notation cannot be said to constitute a phonetic analysis
of a language, nor can any mere enumeration of the s}Tnbols employed,
however precise the accompanying definitions of their acoustic values,
possibly be accounted as such. A true phonetic analysis is provided
only by a systematic presentation of the material investigated on the
basis of alternances established in specific contexts, together with
precise definitions of the acoustic and physiological characteristics
of the sounds symbolized.
Analyses of this type are particularly valuable both to the linguistic
historian and to those engaged in the practical teaching of languages.
To the former the historical study of phonetic phenomena in specific
contexts can alone be profitable. Attempts to investigate data
divorced from the relevant context have repeatedly led, and must
inevitably lead, to bewildering confusion.
The value of systematic phonetic analysis in Hnguistic pedagogy
needs no emphasis. The establishment of alternance tables such as
those presented above is the first task of those who seek to give a truly
linguistic basis to their practical phonetic teaching, since only by a
consideration of the particular circumstances in which varying degrees
of phonetic differentiation occur in the foreign language and in the
mother tongue, is it possible to bring order and proportion into what
is otherwise so apt to degenerate into a meaningless travesty of
pedagogical method.
f
Sakadhuma
By f Jabl Charpentier
T ITERATUEE : Weber, Otnina mid Portenta (1858), p. 363 ;
Naxatra, ii (1861), pp. 272, n., 393 ; Ind. Studien, v, 257, x,
65 ; Haug, Sitzber. Bayer. Akad. Wtss., 1875, ii, 506 ; Zimmer,
Ai. Leben, p. 353 ; Ludwig, Rig-Veda, iii, 187 ; Bloomfield, JAOS.,
xiii, p. cxxxiv sqq. ; AJPh., vii, 484 sqq. ; Atharva Veda, p. 81 ;
SBE., xlii, 532 sqq. ; Whitney-Lauman, Atkarva-Veda, pp. 377 sq. ;
Henry, La Magie dans VInde antique, pp. 68 sqq. ; Oldenberg, RV-
Noten, i, 160 ; Caland, Ai. Zauberritual, pp. 16, n. 13 ; 175, n. 8 ;
Papesso, Inni dell’ Atharva-Veda, p. 127, n.
The not unknown hymn AV., vi, 128,^ runs as follows : — •
sakadhumam naksatraryi yad rdjdnam akurvata \
bhadrdham asmai jrrdyacchan idam vd.strani asdditi || 1 |1
hhadrdham no madhyandine bhadrdham sdyam astu nah |
bhadrdham no ahndm prdtd rdtri bhadrdham astu nah || 2 |1
ahordtrdbhydm naksatrebhyah surydcandramdsdbhydm \
hhadrdham asmahhyam rdjam chakadhuma tvam krdhi |1 3 |1
yo no hhadrdham akarah sdyam naktam aiho diva \
tasmai te naksatrardja sakadhuma sadd namah || 4 ||
“ When the constellations made Sakadhuma their king they
bestowed upon him auspicious time : ‘ this shall be his realm.’ * (1).
“ Auspicious time for us at noon, auspicious time be ours at evening-
tide ; auspicious time for us at the dawn of the days, auspicious time
may night be to us. (2)
“ During day-and-night, during (the time of) the constellations,
during (the time of) sun and moon do thou, 0 king Sakadhuma, make
auspicious time for us.® (3)
^ The verses 1-2 and 4 are found in the Paipjpalada xix, with rather strong
differences in vv. 1-2 ; these varix lectioTies are, however, not important (except
perhaps in v. Id : tato rastram ajdycUa instead of idam rdstram asdd iti of the Vulgate),
* Better perhaps Ppp> : t<ito rdstram ajdyala “ thus did his rule originate ”,
® This verse is partly metrically irregular and is not found in the Ppp., it may
perhaps be a later interpolation. But quite apart from that, Whitney’s translation
(“ from day-and-night ”, etc.) is unintelligible to me. There is, of course, not the
slightest doubt that ahordtrdbhydm may mean “ during day-and-night ” (cf. Delbruck,
Ai» SynXaXy p. 130 ; Speyer, Ved. u. Skt. Syntojc, p. 13), though there may seem to
exist some difficulty concerning the other two words. However, I can see little
difference between, e.g., naksatrebhyah and naksairesu in a connection like this one.
“ The time of the constellations ” would probably be = sdyam, the very beginning
of the night when the stars first become visible.
450
JARL CHARPENTIER —
“ Thou who hast made auspicious time for us at evening-tide,
by night and then by day — hail always to thee, 0 SaJcadhuma, king of
the constellations.” (4)
The passages of the Kausikasutra where this hymn is mentioned,
and where consequently the word sakadJiuma again occurs have been
indicated and discussed by Whitney and Bloomfield and need not be
quoted here. It is also found in the compound sakadhumaja- in *4F.,
viii, 6, 15 : —
yesam pascal prapaddni purah parsmh piiro mukhd |
khalajdh sakadhumaja urunda ye ca nMpnatdh kurnhhamuskd aydsavah |
tdn asyd hrahmanaspate prallbodhena ^ ndsaya |1
In this verse, to which we shall not return in the following, the
sense of sakadhuma- is quite clear. It simply means “ dung-smoke ” ;
and it is a well-known belief that evil spirits sometimes have their
favourite abode in dung-hills.
The word sakadhuma' does not seem to occur otherwhere except
in Katyayana's Sarvdnukramani, p. 11 ; there it is simply an
explanation of the sakmnayo dhumah, which occurs in the brahnwdya,
RV., i, 164, 43 (= AV., ix, 10, 25)
sakamdyam dhumam drdd apasyatn visuvdtd para endvarena [
uksdnam prsniiu apacanta vlras tdni dhdrmdyii prathamdny dsan |1
The formal translation of this verse is not a difficult one, with the
exception of the words visuvdtd para endvarena.^ Anyhow, the sense
must be something like this ; From far off I beheld the smoke of
dung in the middle room (or : on the middlemost day) further off
from this lower one ; men cooked a speckled bull — these were the
first settlements.” ® Whatever the sense is — and we shall revert to
that presently — there can be no doubt whatsoever that sakatnayo
^ pratibodhen42y although approved of by the dictionaries, can scarcely be quite
correct. Perhaps pratlbadhena, which has undoubtedly been suggested by some scholar,
may help us (other suggestions are rather too far ofif).
^ Henry, MSL., ix, 247, translates : “ entre ciel et terre ”, which is scarcely possible.
Geldner, Rigveda^ i, 212, has: '"Sn der Mitte jenseits dieses unteren (Raumes),” which
is formally quite acceptable though the sense remains somewhat obscure. Grassmann,
Wb.^ 1308 supplies agnina and seems to think that we should translate “ by the
middle (fire) which is out of the question. Ludwig, RV ii, 583, is, as usual,
extremely obscure.
® The words tani dluirnidni praihamdni dsan, which recur in i, 164, 50 (-= x, 90, 16),
must mean something like this ; dharman-i, according to my humble opinion, does not
here mean “ laws ” or “ customs
^AKADHUMA
451
dhumah is here = sakadhumah, and that, consequently, it means
some sort of star or constellation. We have here a piece of old celestial
lore,i and we shall presently try to find out whether the real sense of
the verse can by any means be got at.
Great authorities have judged quite differently concerning the
real sense of the word saJcadhiima. Thus Weber, Omina und Portcnta,
p. 363, believed it to mean the fire lit before the break of the dawn
while the stars are still visible ; by the rising and falling of its smoke
it would prognosticate the weather of the day to come. To Weber
this salcadhiima appeared anyhow to be identical with the sakamayo
dhumah of the J?F., i, 164, 43. Haug, however, pointed out that this
passage does undoubtedly speak of the sacrifice of a bull, and that the
sacrificial fire is fed by wood and not by cow-dung.- Caland again,®
thinks that sakadhutna is originally a lump of cow-dung, which
contains the smoke within itself; and in AV., vi, 128, where he is
called the “ king of the constellations ” the sakadhuma is in reality
Agni in his aspect of the moon, the adhipali of the naksatras.^
Henry ® believes that uksd prsnih can mean nothing but a storm-
cloud, and that /?F., i, 164, 43, describes the preparation of the breaking
of a thunderstorm. This, although rather eccentric, tallies fairly well
with the explanation of Sayana, which is, however, not decisive,
although an authority like Geldner describes it as possible.®
Already Roth was of the opinion that iakadhuma must needs mean
a constellation, though I cannot find that he especially pointed at the
Milkj' Way. A similar opinion has been more or less \d\’idly endorsed
by Whitney, by Ludwig, by Zimmer, by Oldenberg, and by Macdonell
and Keith.’ Most of these authorities speak of no special asterism
though there are hints that the name would perhaps most probably
designate that giant cluster of solar systems called the iVlilky Way.
Of that suggestion we shall have to say a few words presently.
Finally we have to mention the opinion of Bloomfield, which stands
^ Cf. Oldenberg, BV.-Xoten, i, 160.
* The reasoning of Ludwig, Bigveda, iv, 456, concerning this detail appears to be
rather a casuistic one.
* Zauberritualy p. 16, n. 13 ; p. 175, n. 8.
* Cf. V, 24, 10 : vi, 86, 2 ; m, iii, 4, 5, 1 {= Paraskara, i, 5) ; PBr. 5, 9.
® MSL,t ix, 247 {cf. La Magie dans Vlnde antiquey pp. 68 sqq.),
® Geldner himself would not like to decide whether sakadhuma does really mean
a constellation, dung-smoke or (according to Bloomfield) a weather-prophet.
“ Vedic IndeXy ii, 346.
452
JARL CHASPENTIER — •
all by itself.^ His case may be shortly stated as follows. Darila, in
commenting upon the paribkdsdsutra Kaus. 8, 17 ; pramandosira-
salalyupadhdnasakadhunm jarantah “ (when in the following are
mentioned) the plants pramanda and usira, a porcupine’s bristle,^
a pillow, and a salcadhuma, they are old ones ”, interprets salcadhuma
by hrah'tmrM. This, of course, is entirely without value as it seems
absolutely clear that salcadhuma means simply a smoking lump of
dung {sakrtpinda) ; and such a one must be old, i.e. dry. Just as
valueless is the Atharvanlya-paddhati on Kaus. 76, 19-20, where it is
said that at the wedding four sakadhumah recite the suryapdtha,
as it seems perfectly clear that the author had not the sbghtest idea
what is meant by sakadhmna. Finally, in Kaus. 50, 13-16, a ceremony
for prognosticating fair weather is prescribed, in which lumps of
dung — not burning ones, though — are placed on the linibs of an old
Brah min who seems to be styled sakadhuma.^ Summing up these
passages Bloomfield concluded that sakadhuma could mean nothing
but “ weather-prophet ”, a suggestion that has met with but scant
applause. In spite of Bloomfield’s undoubted authority in
Atharvavedicis there is not the slightest reason to assume that an
earthly weather-prophet should be styled raja naksatrdwm as is the
case in vi, 128, 1, 4 ; and thus the whole suggestion seems to me to
fall flat.
^aka-dhwm- “ dung-smoke ” is a compound, the first part of which
consists of saka(n)-, the heteroclitic stem interchanging with
iakrt.* According to Wackernagel, Ai. Gramm., ii, 1, 56, older
^ The passages in which Bloomfield has dealt with ialcadhiima have already been
quoted above.
^ Why Bloomfield tranalates salalt by “ a boar’s bristle ” is not clear to me, the
ialalt being well known from the stmarUonnayaJia.
^ This may be deduced from 50, 16, but is by no means quite clear.
* The etymology of sakan- now generally accepted seems to be the one
connecting it with Konpos “ dung, manure, filth ”, and which is of rather recent date
(cf. Bartholomae, BB., sv, 41 ; Fick, VergL Wb.,* i, 421 ; Kretschmer, Einleitung,
p. 353). T^et me admit at once that this etymology appears to me entirely inadmissible.
The correct etymology of Konpos is found already in Pott, Et. Forsch.,^ ii, 205 sq. ;
Benfey, Griech. WurzeUex, i, 269, etc. ; the word belongs to Kanvos “ smoke, fume”,
Kairooj to breathe forth ”, €yKd‘rrr€i' €K7rv€Z Hes. ; lit. kvdpas fume, odour ,
kvepiii, kvepti “ to breathe forth, to smoke ” ; Slav, kopu “ fume ”, etc. Lat. vapor
does not belong here, but to ONorse vafr-lozi “ a magic wall of flames surrounding
Asgard ”, vdfudr, name of the wind, vdfa “ to totter, to shake ”, Germ, ivabem, cf.
Danielsson, Gramm, anmarkningar^ i, 16, n. 2; Johansson, Goett. gel. Anz.y 1890,
pp. 767 sq. ; BB.i xviii, 31. Armen, k^amt “ wind ” (Liden, Armen. Stud., pp. 124 sq.)
scarcely belongs here. The imdoubted relationship between ha-t. fumus ajad fimus well
illustrates the connection between Karivos and Koirpos.
SAKADHUMA
463
compounds such as saka-piUa- ^ N.pr., saka-pinda- “ lump of dung ”,
VS., iakaidJia- “ dung-fire ” have only saka(n)- as their first member,
while later ones, from the times of the sutras on (cf. Panini, iii, 2, 24),
have sakrt°, cf., e.g., sakrt-pinda- (Apastamba), sakrd-nti- “ dung-
flow” (Hiranyakesin, Grhyasutra, i, 16, 9, according to Caland, ZDMG.,
liii, 214). The parallel yakrl, yaka(n)- “ liver ”, appears only in
compounds of a quite late date with yakrt° as their first member,
such as yakrl-lmna(n)- N.pr. of a people (MBh.).^ Greek compounds
with rjvaTo- as their first member (e.g. r^naTo-aKOTreo}, etc.) are
of very late date and furnish us with no allusion to the original status
of compounds formed from heteroclitica like yakrt and sakp.. The
Latin stems jecur-, jecin{or)- also carry us no further as far as com-
position is concerned.
There can thus be no doubt at all concerning either the formation
or the meaning of sakadhuma. Its original sense is “ dung-smoke ”,
“ smoke of burning (lumps of) dung.” This, however, will not do for
AV., vi, 128, where sakadhurm is styled “ king of the constellations ”.
To find out its sense in this passage we must return to the suggestions
of older authorities, viz. that sakadhuma is the name of a constellation.
The question is only this, whether we can find out what special con-
stellation was called in Vedic times “ dung-smoke ”.
There have been put forth mild suggestions that the celestial
phenomenon called sakadhuma should in reahty be the Milky Way.
Amongst the many different interpretations by which various people —
Greeks, Hindus, Chinese, Negroes, Indians, etc. — have tried to solve
the riddle of this most prominent ornament of the nocturnal sky there
is scarcely even an allusion to its being a column of smoke.® Generally
the Milky Way is believed to be a path or an enormous river stretching
• right across the heavens ; however, even if there be also a great number
^ This Sakaputa is met with in the RV.^ x, 132, 5 : asmin sv etdc chdkaputa eno
kite mitre nxgatdn hanti vtrdn, within a hymn which is more or less entirely obscure
(cf. Ludwig, Rigveda, iv, 123 sq. ; Oldenberg, RV,-Noteny ii, 349 sq.). There is,
however, no doubt at all that this is the Sakaputa spoken of in Jaim. Br.y i, 171 (Caland,
Verhandel. Akad. Amsterdam Afd. Letterkunde, xix, 4 (1919), 66 sq.), and that with
the help of this passage the hymn may be partly reconstructed, though no such attempt
can, of course, be made here.
* What Brugmann, Grundrissy* ii, 1, 581, says concerning these compounds is
incomplete and insufficient.
* In the Symbolse philologicse 0,A. Danielsson dicatae (Uppsala, 1932), pp. 13-42.
I have collected somewhat ample materials concerning the lore of the Milky W^ay
amongst various peoples. This collection was meant to illuminate the puzzling Homeric
expression {iv) vvktos dpoXyw, which is probably itself an old poetic name of the
Milky Way.
454
JABL CH^KPENTIER — -
of other ideas attached to it the one of its consisting of smoke is either
entirely unknown or, at least, confined to some very remote and
isolated parts of the earthd It may be taken for absolutely granted
that it is not kno%vn within India.
There is, however, a constellation which might perhaps quite well
be called the " king of the naksatras ”, viz the krttikdh, the Pleiades. It
is quite irrelevant at what time the krttikdh became the constellation
of the vernal equinox, whether at 2500-2300 B.c. or perhaps some
centuries earlier or later." For oiu" present purpose it is quite sufficient
that during Yedic times in general the Krttikdh are considered to be the
first among the constellations of the old lunar zodiac and might
thus well be looked upon as having been installed as their sovereign
hy the other naksatras. The one objection to such a suggestion is
that the names of the consteUations are generally feminine ones ;
and it would certainly be contrary to Indian ideas if the asterisms
were to be ruled by a female being. Such an objection, however, falls
flat if we suppose that the name was at a certain period not krttikd{h)
but sakadhumah, which is, of course, of masculine gender and presents
no such difficulty.
If we suppose that sakadhunia is really an older name of the
Pleiades we shall perhaps obtain a somewhat plausible explanation of
that puzzling verse, RV., i, 164, 43, quoted above. From far off,”
it runs, “ I beheld the smoke of dung . . . ® men cooked a speckled
bull — these were the first settlements.” We must remember, I venture
to think, that the Pleiades are closely attached to the constellation
called the Bull {rrsabha). are in reality the seven stars tj, etc., Tauri.
Immediately in front of them is the sign of RohinI (a, y, S, e. Tauri)
which is generally compared with a wagon (mkata).* There is absolutely
no proof that the Hindus of yore looked upon this constellation as
being the celestial bullock ® ; but on the other hand there is also no
^ As far as I am aware only the aborigines of certain parts of Australia consider the
Milky Way to consist of smoke, viz. the smoke from the camp-fires of dead warriors.
Cf. Stanbridge, Trans, Ethnol. Sor. X.S.^ i (1861), 302.
^ Cf. Weber, hid, Sind., ii, 240, 413 .sq. ; Ind. Skizzen^ pp. 97 sq., 136 ; Xaxatrn,
ii, 364 ; Jacobi. FaMgr. Foih,, p. 71 : etc. The words of Weber Naxatra^ ii, 277, con-
cerning the date of the substitution of the krttikdh for rohirii may still be quoted as
a beneficent criticism of the somewhat extravagant theories of Professor Jacobi.
® I leave out here the words visuvdfd para enuvarena, which are somewhat obscure
and cannot possibly be of any decisive importance in this connection.
* Cf., e.g., Kirfel, Kosmographie der Inder^ p. 138.
^ In the Vedas the stars are sometimes styled “ bulls ’* {usrdh, gdvah), cf. Symh.
phil, 0. A. Danielsson dicatae, pp. 25 sq.
SAKADHUMA
455
proof that they did not. And I shall thus venture to suggest that the
contents of this puzzling verse is really a scrap of star-lore, according
to which men cooked a bull (the constellation of Taurus) on a dung-fire,
the smoke of which, the “ dung-smoke is in reality the Pleiades. ^
Though I have collected somewhat extensive materials concerning
the names and lore of the Pleiades amongst various peoples. I am not
aware of any special instances where they are looked upon as being
a sort of heavenly smoke. This in itself is, however, scarcely a valid
objection against the suggestion put forth above. For undoubtedly
the Pleiades are sometimes believed to be a light-coloured cloud in the
sky ; and there is little difference between *’ cloud and “ smoke
especially as the “ dung-smoke ” (sakadhiima) is of a whiti.sh colour.
However, there seems to exist within India at least one proof of the
fact that the Pleiades were sometimes connected with the idea of smoke.
Hemacandra, Desln, 5, 62, has preserved an expression of probably
popular origin, viz. dhuimddhayanmhmo krtlihah. With this word
seem to be connected dhutnaddhao tataka niahisasca in 5, 63, and
dhuma'tnahisl , which in 5, 61, together with three s}'nonyms.® is
interpreted by nlhdra “ a cloud The Pleiades consequently were
also known as the dkumadkvajamahisyah. The pond (tataka) with the
mist floating over its surface is dhumadhvaja ; there is. however, a
difiiculty in understanding how the buffalo (makisa) could also be called
dhumadhvaja, unless such an explanation is simply a lapsus caused
by the misunderstanding of the dhumadhvajamahifi. As far as I
understand the word mahisl in this compound cannot be the usual
one meaning “ a female buffalo ” ; it must be another, otherwise
unknown word meaning “ mist, fog, cloud which is connected with
p. mahikd “ mist, cloud ” and its possible relationship.* It even seems
doubtful to me whether mahisl, mahisl, is the correct and original
7.7 ^
^ The last words of the verse : tdni dhdrmani prafhamdny dsan I have ventured
to translate ; “ these were the first settlements/’ This, according to ray humble
opinion, means that the men of yore were at their death metamorphosized into stars
and thus had their abode in the nocturnal sky, an idea which is by no means foreign to
Ancient India (cf., e.g.. Schermann, Am Vrquelly vi, 5 sqq., and the present writer,
Symb. phil. 0. A. JJanielsson dicataCy pp. 30 sq.). According to such an interpretation
the deiah in i?!’., i, 164, 50 (= x, 90, 16) are also called the first settlements ”,
which is, of course, by no means impossible.
^ Cf., e.g., Meghadiita, 68.
^ Viz. dhumariy dhumid, and dhumasiha. According to the Petersburg Dict.y
dhumamahisi in certain native dictionaries means “ a cloud On dhumarl cf.
Zacharia, Beifrdge, pp. 55, 66, 85 ; dhumarikd occurs in the Comm, on Gaudavaha, 639.
* Cf. ZDMG.y Ixxiii, 146 sqq., and J. J. Meyer, Hindu Tales, p. 112, n.
456
SAKADHUMA
form of this word, but this must be left aside here. Be it enough to
state that the name dhumadkvajamahisyah may probably denote the
Pleiades as “ the smoky cloud ” or something very much like that.^
That the Pleiades rule over the changes of the weather is a well-
known idea amongst many peoples ^ ; and many are the peoples all
over the earth who begin their year either with the rise or with the
setting of the Pleiades. As, however, I shall hope to deal with these
topics in the near future, I cannot enter upon them here. We may only
remember that amongst the names of these seven stars preserved to
us in the Yajur-vedas {ambd, duld, nitatrii, abhrayanfi, meghayanti,
varsayantT, cupunikd),^ there are at least three {abhrayanti,
meghayanti, varsayantl) which allude to their domination over the
clouds and rains. To pray to them for fair weather, to try to
prognosticate from their appearance the weather of the day to come
would only be quite a natural thing to do. This is what I think is
included in the Atharvanic hymn to the salcadhuma and in the
ceremonies performed with the help of it.
1 The krttikdb, as is well known, are also called bahuldh (cf. Panini, iv, 3, 34 ;
a person bom under this constellation is called Buhula). This scarcely means “ the
numerous ones ” (thus Weber, Naxatra, ii, 368), but rather “ the obscure, dark ones ”
or possibly “ the ones crammed together ”.
* Cf., e.g., the quotations from Aratus and Moiro in Athensms, xi, 490 A ; 491
B, etc.
’ TS; iv, 4, 5, 1 ; Kdth., xl, 4 ; MS., ii, 8, 13 (with the variants bula—a, simple
mistake — and stanayanti instead of varsayanti) ; cf. TBr., iii, 1, 4, 1.
Purana Legends and the Prakrit Tradition in New
Indo- Aryan
By SuNiTi Kumar Chatteeji
TT is now generally admitted that a great deal of the ancient and
J- medieval myth and legend enshrined in the Sanskrit epics and
Puranas is of non-Aryan origin, and that even in Vedic m\'thology
certain pre-Aryan elements are present. Puranic myths of the gods
and legends of kings, heroes, and sages, in the form in which we find
them in the Sanskrit works, represent undoubtedly a considerable
amount of modification from their original forms, whether Aryan or
non- Aryan : witness, e.g., the treatment of what would appear to be
a genuine Aryan (? Indo-European) saga — that of Pururavas and
Urvasi, as we find it in the Rigveda and the Satapatha Brahmana, in
the Visnu Purana and in the subsequent Puranas. The non-Aryan
speaking masses in Northern India became Aryanized in language, and
their tales and legends were retold as a matter of course in the Aryan
language of their adoption. A good many of these were Sanskritized
and obtained a place in the Puranas and other Brahmanical literature,
and so received the stamp of official or orthodox Hinduism. These
Hinduized stories in their Sanskrit version then obtained a prestige
far above that of the older or original versions, which might linger
on for some time, but they were inevitably lost with the loss of the
non-Aryan language and non-Aryan tradition, or, when they were of
Aryan origin, with the later rehandling by Brahman scholasticism.
The Aryanized or Hindu version in Sanskrit, if it came into being
sufficiently early, imposed the Aryanized or Sanskrit names of the
gods and heroes among the people, though here and there some debased
non- Aryan names might continue in connection with village cults and
ceremonial which were tolerated, and at times even encouraged with
good humoured indulgence, by the followers of the official Brahmanical
rehgion.
AVhen the Sanskrit names were in use early, before the close of the
Middle Indo-Aryan (Prakrit) period, we might expect to find Prakrit
forms of these names to be in current use among the people as something
received traditionally, like the religion and like the general mass of
words in the language itself. And such MIA. names would in their turn
VOL. vra. PARTS 2 .CND 3. 30
458
S. K. CHATTEEJI —
be expected to be handed down, together with the m\dh and legend
and the religions as well as linguistic traditions, to the Xew Indo-
Aryan (vernacular) speeches. But in this matter, more than in any
other thing, there has been a constant interference from the sacred
language. Sanskrit. This interference began with the beginning of
MIA. : thus the MIA. hamhana. hambhann. hdhhhana < hrdhnutna has
continued down to NIA. (e.g. Bengali bdnnii), Bihari hdhhcnt. etc.),
but aheady in Pali the vernacular word is ousted by the Sanskrit
brdhnana.
The Sanskrit versions of the Purana legends became the standard
versions. They became pan-Indian during the la.st 2,000 or 1.500 years.
A local cult or a local deity with its local legends could address itself
to an all-India audience only through a Sanskrit Purana in which it
found a place, either as an independent cult or by incorporation into
or s^Tithesis with some other cult or deity. In this way it became one
of uniyersal Hindu appeal and universal Hindu popularity, although
originally it was not much known beyond its own little tribe or
district.
The Rama story, the Mahabharata story, the Krsna legend —
all these, and many more, are now so very much the common property
of the Indian (Hindu) people that it is hard to realize that 2.000 or
2.500 years ago a good many of them were either not in existence
(at least in the form in which they are current now) or were not known
to the greater part of the Indian world of the day. They were passed
on from one part of the country to another as a part of the Hindu
(Brahmanical, Jaina, and Buddhist) religion which seems to have
been taking shape in the Upper Ganges Valley, through an initial
fusion of the Aryan with the Dra\-idian and Austria peoples and
cultures, during tlie middle of the first millennium B.c. From the
Pali Jiitaka it would appear that even during the centuries immediately
before Christ the Rama story had not yet crystallized — it was still
in a fluid state, and that the Mahabharata story was not yet fully
known to the Aryan and Arvanized people of the East, among whom
the Buddhist canon, including the Jataka (the latter at least in its
primitive form), grew. About the Krsna legend, too, we find versions
and references in the Jataka quite confusing in their contradictions
to and occasional agreements with the Purana and the Mahabharata.
The Brahmanical Ramayana, Mahabharata, and Puranas made a
clean sweep of the older versions, seeminglv vdping them out of the
popular mind (I say seemingly because in some cases these extra-
PURAXA LEGEXDS 459
Purana versions persisted, despite the supreme position of the Piirana) :
and with the establishment of tlie above works tlie Sanskrit forms of
the names put out of use the vernacular or Prakrit ones, which had
been current in the land from the earlier pre- Aryan or pre-Brahmanical
tradition, or had obtained currency from the stories boiii^ carried
by vernacular oral tradition from province to province duriiifi the
iMIA. period.
We have no means of knowin*; how and when what may be called
“ the Sanskrit Purana tradition " established itself among the people
in a particular locality, but there is no doubt that it was pretty late
in being introduced in some parts of the country. Sanskrit names now
reign supreme in most of the XIA. speeches in either their pure
(talsama) or modified {semi-tatuntna) forms. This can give us no clue
for finding flie likelv period of inf reduction of a Purana legend
among the people of a particular locality — of its coming into vogue
among them. The masses followed their gmmija-dhunna, their village
cults, which could be described as a sort of half-cooked animism on
its way to transformation into higher Hinduism, to which it had in
a vague way affiliated itself. The ffrdnut-dn-afd-^! or village gods and
godlings had their own names, vernacular Indo-Aryan or pre-Aryan,
unmeaning of an}i:hing to their worshippers. The Sanskrit gods and
heroes with their Sanskrit names brought in a new world of dignity
and romance and also of spiritual aspiration, and the grama-deivtds
as a result were cast into obliffion or were identified with the new
deities with their Brahmanical prestige. In a case like this the cultural
status and the power of assimilation and resistance of the people
being brought under the influence of official Brahmanism decided
how far the new pantheon and the new names would be victorious.
Thus in the Dravidian South, in the Tamil land, the Indo-Aryan
Hindu gods imposed their worship and their legends and also their
names (and at the same time they absorbed a great deal from the local
cults and local legends as well), but some of their Dravidian counter-
parts or protot}'pes did not wholly surrender — their native names
have survived ; for we have the Tamil Murukrut beside Kumdra or
Snbmfima>n/a. Mai beside Tbsnu. Korrami beside Dargn. Mdydn
beside Kr.pja. Vfdigdn or Vellaiyon beside Baladeva , and the names
Sira and Snmbhu themselves have been suggested as being early
Sanskritizations from the Dravidian. A study of the onomastics of
the North Indian grdma-devatas (full lists and studies of the cults of
these are still lacking) will undoubtedly reveal as much of interesting
460
S. K. CHATTERJI —
and important facts about pre-Puranic and pre-Hindu religion in
India as that of the Sanskrit mythological and legendary names.
The present question, however, is that of the introduction of
Puranic myths and legends of gods and heroes in the various Indo-
Aryan tracts, and of their original or earher forms. If in this connection
we can find in a NIA. language a name of Prakrit origin, either still
current or obsolescent or wholly' obsolete, used side by side with or
suppressed in recent years hy the corresponding Sanskrit one, that
would certainly be a very strong argument for the assumption that the
story of the particular god or hero in some form or other had become
popular in the MIA. or pre-vernacular period. In this way we may
obtain from Indo-Aryan (and other) linguistics some valuable ancillary
help for the reconstruction of the religious history of Aryan India.
It is, however, necessary to guard against an argumentum ah silentio
in a situation like this, as there is every HkeUhood of an old Prakrit
name having been supplanted by its Sanskrit form. Thus we see from
Al-BirunI how during the last 800 or 900 years some common
geographical names with a rehgious association have become altered
from Prakrit to Sanskrit — old Prakritic names like Mdhura (or more
correctly Mahura as given by Al-BirunI’s elder contemporary Al-‘utbl)
and Jawn, which Al-Blruni heard and wrote down about 1000 A.c. have
now been supplanted by the semi-tatsama forms Mathrd and Jamnd
(Muttra, Jumna = Mathura, Yamuna). But the positive fact presented
by the Prakritic or MIA. vernacular names cannot be ignored, that the
Puranic legend in question actually existed as a popular one, not
confined to the Sanskrit texts, at some period before the development
of the NIA. vernaculars.
Thus, among the Bengali-speaking people, the persistent use
even at the present day, of a number of names of MIA. origin in
connection with the Kmna legend would testify to the fact of its
popularity in the pre-Bengali period, and the occurrence of one or two
names in their Prakritic forms would also point to the existence of
certain episodes or versions of that legend before 1000 A.c., the point
de depart for the NIA. vernaculars. The same thing we cannot say'
of the Rama legend : although no MIA. forms of names of characters
from this cycle are found in Bengali, it might have been for aught we
know equally popular with the Krsna legend in the pre-vernacular
period : and, as a matter of fact, there is other emdence pointing to
the existence of old popular versions of the Rama story side by side
with or to the exclusion of the Sanskrit version as in Valmiki. The
PUEAKA LEGENDS
461
following names from the Krsna legend are to be considered : Krsna :
MIA. Kanha, Old Bengali Kdnka, Middle Bengali Kdnha, Kdnq,
Kdnha, Kdna, New Bengali Kdnu, Kdndi (< Kdna + -u, -di).
RddJid, Rddhikd : MIA. Edhid, Rdhia, MB. Rdhi, Rdi, NB. Rdi.
Nanda : MB. Ndnda = Nd^da (with reduced nasal ; cf. Origin
and Development of the Bengali Language, pp. 360-2), Ndda (^rlkrsna-
klrttana).
Kamsa : MB. 0rlkrsna-klrttana) Kdsa. On the basis of this MB.
form, with -3- for -dm-, the Mohammadan historians wrote down
in their Persian histories the name of the Hindu feudatory prince
(who during the second decade of the fifteenth century wrested the
kingdom of Bengal from its Mohammadan ruling house) with an alif
(= long -d-) in the middle, as k’ns = kdns for *kds.
Abhinianyu : MIA. Ahimannu, Ahivannu), Late MIA. *Ahimanna
[Ahiwanna), *Ahivanna, early MB. Aihana (as in the $KK., from
Ahivannu), Aimana (pronounced Aivcana, as in the Harivarnsa of
Bhavananda : iTom Ahimannu), late MB. Aydn, NB. Aydn. (A'iham
appears to be an early West Bengali, and Aimana an early East
Bengali form).
The Bengali forms based on those of MIA. would certainly demon-
strate a continuity of the tradition among the BengaU-speaking people
in the matter of the Krsna legend, at least from the time when
Ahimannu- Ahivannu and Rdhid were current in popular speech.
The form Aihana-Almana-Aydn = Abhimanyu has an important
bearing on the development of the Krsna legend in Bengal.
The episode of the hero’s amours with the gopis in the Krsna
legend grew with the centuries. From the scanty references to t his
Vraja-lild, or sports in Vraja of Krsna and the cowherd maidens, in
the older Puranas like the Visnu Purana (and the Harivarnsa), it
was elaborated by successive generations of poets and Yaisnava
teachers, the process continuing down to the eighteenth century in
Bengal. Mr. Sukumar Sen in his recent work, A History of Brajabuli
Literature (Calcutta University, 1935). which is a very valuable
historical and literary study of their lyric literature of Bengal
Vaishna\dsm, has given a useful conspectus of the “ Historical
Development of the Eadha-Krsna Legend ” (chap, xxiii, pp. 472-
481 ; also chap. xxiv). It seems that the Eadha-Krsna legend,
particularly in a detailed account of the loves of the two divine lovers,
received a greater elaboration in Bengal than anywhere else, certain
matters and episodes not mentioned in the Puranas both early and
462
S. K. CHATTEEJI —
late being found in Bengali Yaislinavite literature only and in no other
vernacular literature. Eadha, as the favourite of Krsna among the
gopis, came to be recognized in the Puranas very late — the Harivarnsa,
the Visnu-Purana, and even the Bhagavata Purana do not knov her
even by name ; on the other hand, elaborate Eadha stories vith her
family tree and entourage all complete and her special status and
mystic place in the Krsna-gopa or Yraja-lTla episode form a fully
established feature in the very late Puranas. the Padma and the
Brahma-vaivarta. the latter of which was compiled in West Bengal
as late as the sixteenth century (vide Eai Bahadur Yogesh Chandra
Yidyanidhi's Bengali articles on the subject). In Bengali Yaishnavism,
Eadha has been exalted even above Krsna ; and while the earlier
Puranas are silent about Eadha, and while in the North Indian
tradition, as e.g. in Sura-dasa, the Braj poet of the sixteenth century,
Eadha. is just an unmarried gopl, the embodiment of the human soul
in quest of God through faith and love, the Bengali tradition, from the
poet Ba(Iu Candidasa, the earliest Bengali writer on the Krspa legend
(fourteenth century (1). fifteenth century (0) onwards, makes Eadha
a married woman, the wife of one Aihana (Aimana, Ayan).
This name in its Sanskrit form is unknown to Bengali ; the Sanskrit
equivalent Abhimanyu occurs for the first time in the Sanskrit literature
of the Bengal school of Yaishnavism during the sixteenth century :
the Brahma-vaivarta Purana gives the name in the form Rdyana,
which is olndously based on the si.xteenth century Bengali Aydn —
and in this way this Purana betrays its indebtedness to sixteenth
century vernacular Bengali sources for this episode. Aydn-Aihana-
Aimana goes back to MI.V. times, the MIA. Ahivannu-Ahimannu
alone can explain the Bengali forms by linking them up with
Ahhhnrmyu : the Bengali forms as derived from MIA. unquestionably
e.stablish an old tradition, and if the name Abhimanyu were a new
one in the Krsna legend (as many names in the Radha-Krsna cycle
are), cropping up or gaining currency at the hand of the Yaisnava
scholars in the si.xteenth century, we shoidd have found only the
tatsama form or a semi-iatmma ecjuivalent for it in Bengali, and no
tadhhava or Prakritic form like Aydn-A'ihana-A'imann.
Eadha's position as the wife of a respectable man who is carried
by the frenzy of her passion for God incarnate as Krsna, quite against
social conventions and social morality and reckless of social obloquy,
gave rise in Bengal Yaishnavism to the doctrine of the Paraklyd or
Another's Wife (Vay, in which the love of man for God could be
PURANA LEGENDS
463
compared only to the intense passion of a married wife for another
man, a passion which may lead her to sacrifice eveiA-thing in the world.
The germs of the Parakiya idea could only be based on the initial
belief that Radha was already a married woman before she loved
Krsna : and it may be suspected that the Parakiya idea arose partly
from the necessity to explain or justify what pained most people
as something anti-social and immoral in a popular legend when the
latter was made the basis of religion and philosophy. Moreover, to
ease the conscience of the Yaisnava faithful, by showing the immaculate
nature of Eadha's conjugal life, her husband Ayiin (Abhimanjm) was
described as an impotent person — -there being some insistence on
this point among the later Yaisnava writers. But all this, much less
Eadha's marriage, and even her very exi.stence, is unknown to the
older Sanskrit Puiana tradition ; and the North Indian and other
Indian tradition know Radha to be just a simple cowherd maiden,
one among the gopls. whom Suradasa in his Braj l\Tics marries to
Krsna according to the Gandharva rite. How old can this be in the
Radha-Krsna legend, and where would it liave come into being ?
The oldest literary mention of Radha occurs in the ■' Gatha-sapta-
satT " of Hala, which, although it undoubtedly pre.sents older elements
going back possibly to the first century a.c.. cannot date, in its present
form, from an epoch earlier than the middle of the first millennium
after Christ. But it appears there is plastic evidence in Bengal, dating
from the sixth century a.c.. as to the existence of Radha as the yopl
par excellence, the one specially loved by Krsna. AVe have to mention
the famous Paharpur Stupa bas-reliefs in .stone of the Krsna legend,
comprising the unique slab with a pair of youthful lovers whose divine
nature is indicated by a floral halo round their heads, — the lover,
an ephebe {Kisara) and the girl standing side by side in an affectionate
manner in poses which remind one of the characteristic poses for
standiii" Krsna and Radha in later Bengali art. (The famous Pallava
bas-relief on a heroic scale at Mahabalipuram near Madras, one of the
grandest creations of Indian sculpture, depicting the holding of the
Govardhana Hill by Krsna and dating probably from the seventh
century, seems also to show the figure of Radha to the left of the hero :
but this is slightly later than the small slab at Paharpur, where it
will be allowable to see Radha and Krsna in their oldest plastic
repre.sentation.)
Radha (Rahia) is thus attested from literature of pre-vernacular
times, and plastic art seems also to have figured her from the sixth
464
S. K. CHATTEKJI —
century. The Bengali name Ayan-Aihana-Aimana for her husband,
presupposing the MIA. source-forms would establish the occurrence
of this character in the Radha-Krsna story as current in Bengal
about the same time, if not earlier. Radha as a Paraklyd NdyiM,
therefore, can reasonably be deduced from the Prakrit form of her
husband’s name as having existed from MIA. times — ^from the middle
of the first millennium after Christ, at least : the germs of the spiritual
translation of the Parakiyd idea which was formally adopted as an
important doctrine in Bengal Vaishnavism in the sixteenth century,
therefore, can be taken back so early. This conception might have
originated in Bengal, with the Buddhistic Sahaja-yana ; and it
remained confined to Bengal, and restricted among a local group of
Vaisnavas (among whom we have to count Jayadeva, the author of
the Gitagovinda, of the twelfth century), who drew upon the Radha-
Krsna story as a spiritual pattern, or spiritual experience and exercise,
up to the sixteenth century, when, under the impetus of the Vaisnava
re^fival inaugurated by Caitanya, it asserted itself as an idea, and was
adopted by the followers of Caitanya, who built up the Gautjiya or
Bengal Vaisnava doctrines, and was carried by them to Purl, to
Brindaban, and to other centres.
The Prakritic names current in the vernacular are thus a strong
evidence for the antiquity of some of the features of the Krsna legend,
despite the silence of the early Puranas about them. The Prakrit
tradition in Bengali in this matter can thus be utilized as evidence for
the existence of a base for the Paraklyd-vdda in pre-Mohammadan
times, possibly almost coevally with the Radha story itself — at least
in Bengal.
Later legends in Bengal and in the late Puranas give the names of
the cowherd attendants and associates of Krsna and the gopi friends
of Radha. The Bhagavata Purana knows some of the former, but the
latter are gradually created in Bengal Vaishnavism, eight of them
(asta-sakhi) being finally named and selected as special confidantes
of Radha, and superior to the rest, whole hosts of whose names occur
in the different works. In the earlier versions the gojiis form merely
a crowd, as a background for Radha, and one old woman, the beldame
Baddyi (so she is called in the SKK. and in other MB. works, the name
commonly occurring as Baddi), stands out as Radha's confidante,
her chaperone, who acts as a sort of kuttirii or duti (go-between or
messenger) between her and her lover. In some late Sanskrit works
Baddyi’ s name is given simply as Jarafi or “ the old woman ”. Latterly
PUEANA LEGENDS
465
the character of Badayi took up a Sanskrit name of one of Radha's
Sakhls — she became merged in Vrnda, popularly called “ VrndadutI ’’
in Bengal, who has assumed most of the functions of Badaja. The
Prakritic name Badayi (<bada “big, great”, of uncertain origin,
and dyi < dyi = dryikd “ grandmother ”) may be taken to indicate
that this beldame also comes down from the pre-vernacular tradition
of the legend, and belongs to the older, non-Puranic or pre-Puranic
stratum.
The above Prakritic forms in NIA. Bengali are thus of great value
in studying the origin and growth of the Purana legend of Krsna-
Radha in Bengal. But the same cannot be said of the Rama legend.
Only ts. and sts. forms of the Sanskrit names are in use in Bengal.
In the Awadhi tract, on the other hand, as we find from Tulasidasa's
works, forms of the names of the Ramayana characters which could
only have come down from the MIA. are extant. Thus : —
Slid : MIA. sid, sia, Old Eastern Hindi siya, siya, present day siyd,
sayd (e.g. the Bengali name Saydrdm, sometimes “ corrected ” to
Sahdy-rdm, which is borrowed from Eastern Hindi siydrdm).
Laksmana : MIA. Lakkhana, OEH. Lakhana. The influence of
the Sanskrit form with Ld- was certainly responsible for modifying
the Ld- of an expected *Ldkhana.
Rdma : the form commonly used in NIA. is the Sanskrit Ratna
(Rdm), instead of the expected vernacular *Rdiva, Rdw : but the
fact that in the early Hindi MSS., Rdma is frequently — almost
always — spelt as TTHT = Rdmma, would seem to suggest that this
spelling with the nasal was a compromise between a popular
pronunciation *Rdwa or Rdiva and the learned form Rdrm.
Vasistfia : Hindi Basith. This name has changed its meaning, and
in some respects has degenerated semantically in Hindustani.
No other names bear the stamp of MIA. on them, but these would
enable us to assume that the popularity of the Rama story, in
the Eastern Hindi districts at least, dates from the MIA. period.
The great gods of Hinduism were worshipped in the pre-vernacular
period in Bengal, but the vernacular Prakrit names which were in
existence have in most cases been ousted by Sanskrit tss. or stss. If
a sts. kestd (or kestd) has not been able to make the Prakritic kdnu,
kdndi obsolete, even in present-day Bengali, another sts. Bistu (or
Bistu), nowadays “ corrected ” to Bisnii, in pronunciation, has driven
out the Bengali equivalents of the MIA. Venhu, Vinhu. A MB.
Baftibha occurs as a sts., but now we have the ts. Brahmd only
466
PURANA LEGENDS
(pronounced as Bromha, or Bemhd, Bemmd in folk Bengali). Siva
occurs as a ts. Sibd. Sibo, Sib in Bengali, but in the Bengali surname
Si it is just likely that we have the Prakritic form. (This Si may equally
be from MIA. siha = simha : but sihgJia and not slha seems to haye
been the form natiye to the source — Prakrit of Bengah, slim being
a "Western Indian, probably also a ^Midland, form : cf. Simhapura
= Siitrjur in Bengal (or Sihgur = Sriiiga-pura ?), but in Kathiawad
we haye Sihaur, and in RajasthanT $1 = simha)
Prakritic names of deities which haye sur\uyed in Bengali are
Diigl, a folk-form, often used in contempt, as an equiyalent of the name
Durgd [Durgikd > Duggid > Dtlgl), and Sdth or Sd(h (pron. Sdth)
= the goddess Sasthl. a form of ilatika worshipped on the sixth day
of childbirth {Sasthl > salthl, satthl > sdthl > sdthi > sdth : the
spelling with s- is due to Sanskrit influence). The form sdth is now
mostly used as a pious exclamation from mothers and grandmothers
to ayert a bad omen from little children — it is merely an inyocation
to the goddess Sasthl. Kail in Bengali might be the ts. Kail, but it is
equally like!}’ that it is the tbh. form of Kdlikd.
M hen Bengali and other NIA. languages started on their new career
the number of such traditional Prakrit names was undoubtedly much
greater. But after the first century and a half of the storm of the Turki
conquest, when much of Hindu religion and culture was endangered,
a renaissance of Hindu life began. There were two streams in which
this reyiyed cultural life flowed : the Bhakti movement on the one
hand, and the wlgarization or popularization of the Puranas on the
other. Sanskrit exerted a tremendous influence on both, and
Sanskritization of Prakritic names came in as a matter of course with
the attempts of both the reformers and orthodox Brahmans to
re-establish a higher Hinduism which could stand the onslaught of
Islam both as a political power and as a prosel}'tizing religion.
Establishment of the San.skrit names linked up the Sanskrit Puranas
and the epics with the life of the masses and gave a certain classic
dignity to this aspect of popular Hinduism, but it broke up the Prakrit
tradition or obscured it. The old Prakritic names, precious indications
of the situation in the popular faith in pre-hlohammadan times,
became more and more restricted. The few that still remain or may
be gleaned are well worth investigation, by resuscitation from the older
strata of NIA. literatures and from NIA. folk-speech, for a close
scientific study.
Some New Awromani Material
prepared from the collections of Age Meyer Benedictsen
By Arthur Christexsen'
TN the late Age Meyer Benedictsen's note-books from his \’isit to
Persia and Kurdistan in the year 1901. which are now in my
possession, due to the kindness of Mrs. Meyer Benedictsen, I find some
unpublished Awromani texts dictated to Meyer Benedictsen by
‘Abdu'l-Ghafur, native of Awromand viz. an unfinished tale, not
accompanied by a translation, and the so-called “ Bacmeistersche
Sprachproben I here submit these texts, prepared according to the
principles laid down in Zf.s Diahctes d'Awroindn et de Pdwd
and noted in the system of transcription used in that book.- The
translation of the tale has been made by me with the aid of a few
marginal notes in the manuscript. The paragraphs quoted in the
footnotes refer to the Grammar given in Les DinJectcs d'Awromdn et de
Pdicd. I leave unnoticed a few differences in the forms of the words,
vocalization, etc.
As to the story of Dalla and the Khorasanian, only the beginning
has been noted. Cf. The Wiles of Women, from the Turkish by
J. A. Decourdemanche (London 1928) pp. 77 fi. The principal figure,
Dalla. is, of course, the wily Dallla of Baghdad, known to readers
of the Arabian Xights (vide Chauviu, Biblioyraphie, No. 147).
Abbreviations
Awr. = Benedictsen-Christenscn, Les Dialectes d'Aivromdn, etc.
M.-H. Gur. = 0. Mann, ‘‘ Mundarten der Guran, bearbeitet von
Karl Hadank ” (Kurdisch-Persisehc Forschunyen, hi, 2). Berlin,
1930. Awromani, pp. 367-395.
Soane = “ A Short Anthology of Guran Poetry," by Major E. B.
Soane. JRAS.. 1921, pp. 57 ff.
ar. = Arabic,
awr. = Awromani.
kiind. = Kanduliil.
p. = Persian.
^ Vide Les dialectes d'Aicroman et de Pdud. Textes recueiihs par Ago Meyer
Benedictsen, revus et publics avec des notes et uneesquisse degramraaire par Arthur
Christensen. Cop., 11)21 {Det K<jl. Daiinke V idenskaberm^ St-lskahs historish-filologiske
Meddelelser, vi, 2), p. 3.
- Observe : j is English / is English j.
468
A. CHRISTENSEN —
The Story of Dalla and the Khorasanian
ja sa’r-u-Bay§ajana sa^^sewae be, ismas Xoaja MohammaS be.
Xaila dawlaSman wa ;^aila sahab-;^oSrat ^ be. Zanewas be, farae zarifa
be. A zania lamas porse be ; ta Ina Xoaja MohammaS iraSas karS
balo ziarat-u-janow ;)^oSaj. Be taqSia^-u-;^oSa Aoaja MohammaS
taSarak^-as karS, ja taek^-u-hajajiana lua paj Makka-i-sarffi. Jiira saia
ja Makkana moattal ben bl, ta jara hajas karSa. Cawmaj ® Aoaja
MohammaS amawse pe BaySaj, jawa ® ow jana wes. Dia zanakes
lamas har por-ana,’ na-zajna.® Saras sirr ® be. Cawmaj watas ba
zanakia : “ Ba balma ow law Abu Jafari Helali ramaii.” Dses-u-
Mnes gort, luaj ow law Abu Jafari Helah. Ja lamaw I zane Abu
jafari-san hali karS. Abu jafari tamasaw ramiis karS, watas :
“■ Ai\hc hairan mana.^^ M’atas ; “ Tefl zaruewsen ja lamaw i
zanena, semmse ajab ajal mawo wa qaSamas ;^as na-mawo paj aSaw
babais.” Cawdmaj hurazaj,^® amewa paj jana wesan.
ja-dmaw jara m«ga a zana kanacawas die.^® ASaw babas ja-dmaw
jara inaepa har diia marde.^* Kanacesan da ba daes-u-dajana. Ta
jara saia da jana sotas da pena,*^ namasan nia ba Dala. Dalaw
BaySaj maskur-ana. Cawdmaj janaw babaw wesana nestara.®®
Qor’anas wen§,®* aemmee kanaiiewae sirin-e-mahbub-e-qasagka be.
Har cai mayluq-u-BaySaj be kul asoq ba Dala bie. Dala su-s ba
kaesi na-karS ; ta roiwse piajwae p^orasSni ama ow BaySa. Ja ra-u-
hamamiena yorasani Dalas dia, asoqas bi. Kawt ow sones,®® ta ama
janaw wesan. .Aorasani watas ; “ Aj Dala, bo, su-m-penam b kara.
Dala wat : " ZEgar honarawat bo, aibes man su-t-pena bakfou.
^ Ar.-p. qodrdt. ^ Ar.-p. laqdir. ® Ar.-p. tddaruk. * Awr., p. I 1.6 b.
° = cdweSnmj, idwSmaj, cawdmaj. * § 58. ^ § 105.
“ § 69, cf. § 35, rem. 2 ; here the syllable -nd is affixed to the 3rd person of the
singular.
° The ar.-p. substantive serr used as an adjective.
§ 137 at the end.
Rami, instrument of divination, ordinarily a tray of sand.
§ 124 ; 134 (-i<5).
Adjective used as a substantive (= serr, Benedictsen’s note).
§ 84.
§ 69 ; it .seems to be a special future form.
Ar.-p. ‘ejdl " family ”, “ children.”
1 - § 47. IS § 66 . *s § 46. s" § 61.
§ 129. In Kurdistan, the children are suckled until the age of four or five years
(Beneclictsen’s remark).
§ 128, remark concerning the use of 6 d.
§ 63. ** § 96 ; 1st Preterite : wdndm.
= ruewse. =6 | gj . g 128 .
SOME NEW AWEOMANI MATERIAL
469
Xorasani watas : “ Huraza ! ^ ba balma bazar.” Daes-u-joasan ^
gort u luaj bazar.
QaSera ® g’aiaj * ba bazarana. Sarafesan di, sar-u-dukan-u-
wesow nesta bera ; ;;^ailawa teiaw noqra war-u-dam-u-wesana nia
bera.® .Xbrastei watiis ba Dala : “To cega banisara, ta anion
balu law sarSfi ; inja ® bazana amon ces makaru.” Dala
nistara. Xorasani lua, law sarafi-ow nistara. Siiraf marhaba
karS, watas : “ Dukan-u-wet-an, har ceS lazam-iin buaca.” ’
Xorasani watas : “ Asrafi romi-S ban ? ” Wata.s : “ Boiii, hanam.”
Abrasani watas : “ Kisewa sar ba mor ® boMTa.” ® Saraf lua kisewa
p6r-as aworS ; sar ba mor be. Xorasani watas : “ Luii, kise wsetari
seka-je neisapuri bowra ! ” Saraf lua, kise wsetari sar ba moras
aworS. .Xbrasani watas : “ Lua, kise waetari sekaw irani bowra ! ”
Saraf lua aworSas. Watas : “ Lua, kisewa sekaw iggibsi bowra ! ”
Saraf lua aworSas. Xorasaniwat: “ Lua, kisewa sekaw urusi bowra ! ”
Saraf aworSas. Qaras haeft kisii siir ba mor zaras aworS, waer-u-
d6m^®-u-;^orasani nifcara. Cawdniaj xbrasani jo-jo kisakes wazne
karS, jo-jo niaj§ena ba^alas.^* Saraf wStas : “Ces makari ? ”
Xorasani wat ; “ Luaj malu.” Saraf watas : “ Hseft kisow amon
banierow lua.” Xorasani watas : “ Mal-e-wem-an.” Siirafi watas :
“ Card mal-u-wet-an ? ” Xorasani w'at : “To ;^ejaiat in-ana amon
rut bakarino ? Ci ba zarana ? ” Sarafi watiis dubara : “ Malaw
amon banierow, lua.” Xorasani watas : “ Amon amSna, mo'amela
bakaru ; istse mo'amela n'makaru. SowSa ba zor namawo.”
Ax^r ^ saraf goziaj,^® farae jaktarinsan 2 ® kost,®i ta abl-u-
bazari geler bio, persasan ; “I ma'areka ces-an ? ” SarM
watas ; “ .^i hazarat, i piaja haeft kisa mal-o-amoniis bardan ;
istae namoSoso.” Persasan jii xor^^'Sani ; “ Paj ces nameSajso mal-u-i
1 § 66.
‘ § 55.
’ b-ivacd, § 65.
Ar.-p. al-garaz.
dont = dam.
§ 99.
Cf. jotdrinij jotriniy ^ 123. ^ § 134.
§ 62 ; Pluperfect. ® § 134.
p. sdr be-mohr. ® § 53 : b-dwdrdy bard,
hsejt, but kawt above, p. 468, 1. 22, vide Awr. p. 11, 1. 11.
niaj-s-end. bax^h p- bagal.
p. loxt.
§ 59 ; Subj. 2nd person with the ending -o, probably = ow, § 128, cf. § 131 at
the end.
18 _ 18 § 79.
jdktdrin= jotdritiif p. jdk for awr. jo. § 60 and p. 88, note 2.
22 geler = /aw‘ (Benedictsen’s remark). It is probably the Turkish participle
geJir “ coming
22 p. porsiddn ; guranl pirsdn, Soane, p. 66 ; Kand., M.-H. Gur., p. 188.
2* Ar.-p. 7no'drdkd. § '^3.
2® nd-m^bo-s-o ; ending -o, vide § 131 at the end.
470
A. CHRISTENSEX
piSje 1 Xorasani wat : “ Mal-e-wem-an. Amon anian-a, ino'amelii
biikaru. Cani sarSfi mo'amela namilkaru ; sowSa ba zor namawo.”
Saraf lua sikat o ^ hakimis karS. Hakimi kiasiis - ;:^orasani.
Aorasanisan aworS. Hakimi persas : Ba ci mal-ii-i piaja niima-
Sajno i ” Watas : “ Mal-e-wSiu-an.” Hakimi persas Ja sarafi ;
■ ■ Mal-o-ki-an i hseft kisa ? ‘ ' Sarafi wat : Mal-e-amon-an. " Aorasaiii
wat : '■ Aajr, mal-u-amoii-an." Ahl-u-surSi ^ qararsan da : '' Har
kasi saraf ja j(;orasanisan bar kSma zanas bar kisena can ^ asrafis-ana
cana.® mal-u-58i-an.'' Persasan ja sarafi : ” Halesan bar kisena
cen ^ asrafis-iina canii ? "' Sarafi juab da, watiis : " Xamazanu ! '
Persasan ja p^orasani. Aorasani wat ; " Hiir kisena an nasa®-ana
canii. Waj(;t somarasan karS. bi ziaj u kam qse ^ ;^orasani bie.®
CawSmaj saS dana lirasan® ja sarfifi asa.^® Jii jaza-i-na;;^tiena
ba'azesan da ow sarafi.
CawSmaj ;)(;orasani watas ba Dalii ; Amon cani piajaw-anan ? '
Watiis : " Afiirimd® bariikala ! lajeq-ni amon su-pena-S b'karu.'’
Aorasani das-u-Diilas gort, bardas ow, luaj ow janaw Dala.
Mullajewas talab karS, maharaw wes biiri paj ^orasani. A saewa
luana baxalas, ba wasaias saS be.
CawSmaj Dalii lamas porae bie ja ;^orasani. No msep o no ro si,^®
coar konace ba jarewse die. A coar konace dai,^’ ba das-u-dajana.
Do jara Dala lamas porae bie. No maep o no ro si, jiira knac itaras die.
Ta gawre bie baft lotias aworde, bar kanace wes da ba lotiewiisan.
Haft ziima, baft knace, wes-u-Diilii ciini sus. sapzda nefiiri, nan u
biirgsan garak-an . . .
' Ar.-p. sekajdt ; o is = mo (preposition).
- § 80.
2 Sum, assembly (Benedictsen's remark), perhaps ar.-p. Surd " counsel , “ con-
sultation ”.
’ p. cdnj.
Brepo.sition used as an adverb.
* an ndSd = p. an qdddr (Benedictsen's remark).
■ Cf. § 34.
® The Khora.sanian, having weighed the purses one by one, knows approximately
how much money there is m each.
’ lira, ital. lira.
§ 90.
“ is probably a corrupt form of ar.-p. naqd.
Cf. Awr. p. S6,'l. 20 ; p. 108, 11. 8 and 9.
dfdritn, dfdreiii or dfdrun ? reading uncertain ; ji.afdnn.
mahnr, ar.-p. mahr.
u-dsdt, ar.-p. vasl.
§ 93.
da, § 46.
SOME NEW A^VROMAXi MATERIAL 471
In the city of Baghdad there was a person whose name was Khoja
Mohammad. He M'as very rich and very mighty. He had a wife who
was very handsome. This M'oman was pregnant (lit. : this woman,
her womb was full) ; at last Khoja Mohammad resolved to go on
a pilgrimage to the house of God. By the predestination of God
Khoja Mohammad prepared his provisions [and] with the pilgrims
he went on to the holy Mecca. For three years he dwelt in ilecca,
until he had performed the [ceremonies of] pilgrimage three times.
Then Khoja Mohammad returned to Baghdad and arrived at his house.
He saw that his wife was still pregnant, had not been delivered.
He (lit. ; his head) was struck with astonishment. Then he said to his
wife : “ Come, let us go to Abii Ja'far Hellali the geomancer
(rammal).” He seized the hand of his wife, they went on to Abii
Ja'far Hellah. They gave to Abii Ja'far an account of the pregnancy
(ht. ; the womb) of that woman. Abu Ja'far e.vamined his rami
and said ; " This is (lit. : continues to be) wonderful." He continued ;
“ A child will be born from the womb of this woman, but it will be
a ctu-ious child, and its steps will not be happy for its mother and
father.” Then they get up, came [back] to their own hou.se.
After three months that woman gave [birth to] a girl-child. The
mother and the father, after three months, died both of them. People
gave them into the hands of a nurse. For three years the nurse suckled
them (lit. : gave them milk) ; they gave (lit. ; put on) her the name
Dala. She is known as " Dala of Baghdad ”. Then she remained in the
house of her father. She read the Qur'an, but she was a sweet, gracious,
and nice girl. All created beings who were in Baghdad were in love
with Diila. Dala did not make anyone her husband, until once a day
a man from Khorasan came to Baghdad. On the way to the bath the
Khorasanian saw Diila, fell in love with her. He went (lit. ; fell)
after her, until she came to the house of her family. The Khorasanian
said: “ 0 Dalii, come, make me your husband." Diila said : If you
have skill, there is no fault in that I make you my husband." The
Khorasanian said : " Eise ! let us go to the bazaar." They joined
hands (lit. : took the hand of each other) and went to the bazaar.
For some time they were walking in the bazaar. They saw a money-
changer [who] was sitting in his shop ; he had laid down much gold
and silver before his face. The Khorasanian said to Dala ; “ Sit
down there, whilst I go off to the money-changer ; then you shall
know what I [am able to] do.'’ Dalii sat down. The Khorasanian
went off, sat down with the money-changer. The money-changer
472
A. CHKISTENSEX
bade him welcome, said ; “ This is your own shop, say what you want.’
The Khorasanian said : “ Have you Roman gold-pieces ? ” He said :
Yes, I have. The Khorasanian said : “ Bring [me] a sealed purse.
The money-changer went off, brought him a full purse ; it was sealed.
The Khorasanian said : “ Go [and] bring another purse [full of]
money of Nishapur. ’ The money-changer went off, brought another
sealed purse. The Khorasanian said : “ Go [and] bring another purse
[full of] Iranian money.’’ The money-changer went off, brought it.
[The Khorasanian] said ; “ Go [and] bring a purse [full of] Enghsh
money.” The money-changer went off, brought it. The Khorasanian
said : ‘ Go [and] bring a purse [full of] Russian money.” The money-
changer brought it. In short, he brought seven sealed purses [full of]
gold, laid them before the face of the Khorasanian. After that the
Khorasanian weighed the purses one by one, put them in his pocket
one by one. The money-changer said ; “ What are you doing ? ”
The Khorasanian said : “I am going away.” The money-changer
said : “ Lay down my seven purses and go away.” The Khorasanian
said : “ They are my property.” The money-changer said : “ How
are they your property ? ” The Khorasanian said : “ Do you think
you can rob me ? WTiat is the matter with this gold ? ’ ’ The money-
changer repeated : “ Lay down my property and go away.” The
Khorasanian said : “ It is my trust (i.e. I have entrusted the purses
to you), that I may do business with it ; in such a manner I do not
do business. No profit comes from falsehood.”
Finally the Khorasanian and the money-changer fell to blows,
fought much with each other, until the people of the bazaar came up,
asked : “ What conflict is this ? ” The money-changer said : “Oh
you who are present, this man has seized upon seven purses which
are my property ; now he will not give them back.” They asked the
Khorasanian : “ Why do you not give back the money of that man ? ”
The Khorasanian said ; “ It is my own money. It is my trust [with
which] I shall do business. I do not do business with that money-
changer ; no profit comes from falsehood.” The money-changer
went on, made complaint before the Governor. The Governor
summoned the Khorasanian. They brought the Khorasanian. The
Governor asked : “ Why do you not give back the money of this
man 1 ” He said : “ It is my money.” The Governor asked the money-
changer : “ Whose property are these seven purses ? ” The money-
changer said : “ They are my property.” The Khorasanian said :
No, they are my property.” The people of the assembly proposed
SOME NEW AWROMANI MATERIAL
473
the [following] decision ; “ Anyone [of them], the money-changer
or the Khorasanian, whosoever knows how many gold coins there are
in each purse (lit. : in each purse how many gold coins there are
therein), to him they belong." They asked the money-changer :
‘‘ Now, in each of these purses how many gold coins are there ? ’’
The money-changer gave answer, said : “ I don’t know." They asked
the Khorasanian. The Khorasanian said : “ In each purse there is
so and so much [therein].’’ As they counted their content (lit. : them),
there was what the Khorasanian said, a little more or less. After that
they took a hundred lire from the money-changer [as a fine]. In
return of the cash (?) they dealt some strokes to the money-changer.
Then the Khorasanian said to Dala : “ What kind of man am I ? "
She said : ” Bravo ! May God bless you ! You are worthy [that]
I make you my husband.” The Khorasanian seized the hand of Dala,
took her with him ; they went on to the house of Dala. She sent
for a Molla, entered into marriage with the Khorasanian. That night
she embraced him (lit. ; went into his arms), was happy in the union
with him.
After that Diilii was pregnant by the Khorasanian. Nine months
and nine days passed, she bore four girl-children at the same time.
She gave those four girls in the hands of a nurse. Another time Dalii
was pregnant. Nine months and nine days passed, she bore three
more girl-childreii. As they grew great, she brought seven buffoons,
gave each of her daughters [in marriage] to [one of] the buffoons.
Seven sons-in-law, seven daughters, Dala herself with her husband,
sixteen persons [in all], needed bread and victuals. . . .
Sentences
(Bacmeistersche Sprachproben)
1. AbSa mmmero.
2. Pia kaem ziwo.
3. ASa zaru5 wes wos giirak-
ana.
4. Ja mamaw aSana sot faran.
5. Sus wosas garak-iina.
6. I zana lamiis pora be.
7. Se.s riiwii ciiowaie ^ aSii
kurra bi piiiSa.
8. AS hallai nawo.s-a.
God dies not.
Man lives [but a] short [time].
The mother loves her children.
In the breasts of the mother
there is much milk.
Her husband loves her.
This woman was pregnant.
Six days ago the mother bore
a son.
She is still ill.
VOL. vin. PARTS 2 AND 3.
^ §134.
31
474
A. CHRISTENSEN
9. Kanaca-u-aSi law aSina Her daughter sits weeping at
nisora garawai garawod her side.
10. Zaro garakas niae mzo.^ The child will not suck.
11. Knaca hallai matawo halo The girl-child can not yet walk,
rare.
12. Salewae u diie msege caowaii
aS ama danja.^
13. I coser zaruee gardi salam-
ana.
14. Josa ramaj ramo/ duamin-
sa hur faro, jaraminsa gurania
card, coaraminsa ;;^;oaje ;^o.^
15. I pia kur-a.
16. Zanis kaer-ana.
17. ASa mseznawo wa;^t-e erase
qse karma.
18. Baraw to pasmfii pasmo.*
19. Waia-u-to wmtana.
20. Tatow-samae ;;^abaras biwae.
21. AS kaem ware.
22. Luta delaw luena.
23. Emae diia paj-ma haena ;
haer daesawae p«nj agusae-ma haena.
24. Zoif sar-u-sarej sawz bo.
25. Zawan daSana inaj daem-na-
26. Daes-u-ras-emii maehkam-
taer-ii ja daes-u-caepi.
27. Mu daeraz u barlk-a.
28. Wuni sur-anii.
29. Aswu;^ana saB^t-ana pesaw
taw ana.
30. Masawi caerniis haene, gusas
niaene.
31. I palawara jawas paro.
32. As gnare sar-u-zaminina.
33. A palawara par-e-siawa ja
baias-anane.
* § 99 .
She is born one year and two
months ago.
These four boys (children) are
well, aU of them.
One of them runs, the other
leaps, the third sings, the fourth
laughs.
This man is blind.
His wife is deaf.
She does not hear, when we are
speaking.
Your brother sneezes.
Your sister sleeps.
Your father is awake.
He eats [but] little.
The nose is in the middle of the
face.
We have two feet ; on each
hand we have five fingers.
The hair grows (lit. : turns
green) on the head.
The tongue [and] the teeth
are in the mouth.
Our right hand is stronger than
the left hand.
The hair is long and thin.
The blood is red.
The bones are [as] hard as
stone.
The fish has eyes ; it has not
ears.
This bird flies slowly.
He sits down on the earth.
That bird has black feathers in
its wing[s].
p* be donjd dmdd.
/
2 § 87 .
" § 90 .
SOME NEW AWROMANI MATERIAL
475
34. Dra;^t geiaj-e-sawz iaqaw
gawrgs hiena.
35. I palawara dsenuk-i-tez,
kalicke kotas ban.
36. AS ja halanina beiew caermes
baena.
37. Awir socno, emae wmnie
dukai, beiesae wa zo;^al.
38. Awi ]a ru;)^anana ton wiaro.
39. Maega gawrataer-ina Ja
hasara wa wurditaer-ena Ja ro-
Jawe.
40. Hezi waran ama.
41. Aro sobbanaman caemam
pena kaeft bazar-e-zaerina.
42. Saw tarlk-ana, ro rosni.
43. Emae qse kiirma ba baw-
raml.
44. Tawda smae qse karda ba
bawraml ?
Tbe tree bas green leaves [and]
big branches.
This bird bas a peaky beak
[and] a short tail.
It has white eggs in its nest.
The fire burns, we see the
smoke, the flame, and the coal.
Tbe water of the river flows
quickly.
The moon is greater than the
stars and smaller than the sun.
Yesterday there was a rainfall.
This morning my eyes fell on
a rainbow.
At night-time it is dark, at
day-time it is light.
We speak Awromanl.
Do (can) you speak Awromanl ?
Additions to the Grammar
{Les Dialectes d’Awronmn et de Pdwd, pp. 17-75)
§ 43. Verb substantive. Present sing. 1 andn.
§ 45. “ To go.” Preterite ii, sing. 3 ludna.
§ 46. “ To give.” Present sing. 2 neg. 'udmdhdj, tidmaSdjno ;
3 neg. ndm9?)0. Pret. i, sing. 3 ddi, plur. 3 -sdn dd.
§ 47. “ To come,” also “ to become,” “ to turn.” Pres. sing. 3 :
mawo, neg. ndmawo. Pret. sing. 3 dmd, plur. 3 amewd.
§ 48. “ To see.” Pret. i, plur. 3 -sdn di.
§ 50. “To be able.” Pres. sing. 3 neg. tmtawo.
§ 53. “ To bring.” Imperative sing, bou'rd, Pret. ii, sing. 3 dwdrde.
§ 59. “ To do.” Imperative sing. b’MrC. Pres. sing. 1 tnakdru, neg.
nm&kdrii, 2 mdkdri. Subj. sing. 1, bdkdru, b'kdru, 2 bdkdrino.
§ 61. “ To die.” Pret. i, plur. 3 »«drde.
§ 62. “ To place.” Imperat. sing, bdniero. Pret. sing. 3 nidsdrd,
niajsend. Pluperf. sing. 3 nid herd.
§ 63. “ To sit down.” Imperat. sing, bdnisdrd.
476
SOME MEW AWKOM.USM MATERIAL
§ 64. Pres. sing. 3 hur f9ru “ he leaps ”.
§ 65. ‘"To say.’’ Imperat. sing, buacd.
§ 66. “ To rise.’’ Imperat. sing, hurdza. Pret. plur. 3 hurdzaj.
§ 68. To pass,” " to flow.’" Pres. sing. 3 icidro.
§ 69. “ To be born.” Pret. sing. 3 zajrid. Put. (?) zaruewasn.
§ 70. “To know.” Pres. sing. 1 neg.
§ 73. ■■ To carry off." Pre.s. sing. 3 ban. Pret. sing. 3 bdrddn.
§ 75. Pres. sing. 3 (guranid) cdw ” he sings ".
§ 79. To make war,” " to quarrel. Pret. plur. 3 gozidj.
§ 80. “ To ask for." Pret. i, sing. 3 kiasdk
§ 96. '■ To read." Pret. sing. 3 wene.
§ 98. “ To live.” Pres. sing. 3 ziwo.
Other verbs : —
“ To ask." Pret. sing. 3 persds, plur. 3 persdsdn.
“ To hear.” Pres. sing. 3 neg. msezndwo.
§ 99. luaj mdlu, “ I am going.”
§ 123. Other forms of the Reciprocal Pronoun : jodsdn,
jdJctdrinsdn.
§ 127. Indefinite Pronouns ; jo-jo “ one by one ".
§ 128. Prepositions : o “ to ”, == ow.
§ 134. Adverbs ; cen, con “ how much ", " how many ”. dn
ndsd, “ so (and so) much
Additions to the Vocabulary
(PP-
bdbd “ father ”.
baxal “ armpit ", “ arm ” (p.
bagal).
belesse “ flame ".
Xodrdt ‘‘ power ” (ar.-p. qodrat).
Idqd " branch " (kand. Idq, M.-H.
Our., p. 264).
mdr “ seal ” (p. mohr).
naxt " cash " (?) (p. 7Mqd).
qaras " in short " (ar.-p. al-
garai).
rdniki “ trousers " (coloured).
rojdwe “ sunshine ", “ sun ”
(p. ruz -f db).
120 - 8 )
rut naked ", “ void " (kand. rid,
M.-H. Gur., p. 278 ; p. foxO-
Ukdt “ complaint " (ar.-p. sekd-
jat).
surd " assembly ".
taqSid, “ predestination " (ar.-p.
taqdir).
tez “ sharp " (kand. tizh, M.-H.
Gur., p. 287 ; p. tlz).
zdntd " son-in-law " (kiind. zdmd.
M.-H. Gur., p. 293 ; p. da-
tnM).
zidj “much", “more" (ar.-p.
zidd).
Les nasales en fin de mot en Sanskrit (et latin)
Par A. CuxY
I. Origine des nasales finales
lyT A. MEILLET (MSL. Paris, t. ix. pp. 365-372) enseignait.— depuis
iTX • ji revenu souvent sur la question et s'est toujours proiionce
dans le meme sens — que, a I'accusatif singulier des themes en -o-
du genre anime (inasculin. feminin) et au nominatif-accusatif singuher
des themes en -o- du genre inanime (neutre). la nasale finale etait en
indo-europeen, non pas -m comme on I'ecrit generalement, mais bien -n
(de meme pour les themes eonsonantiques masc.-fem. : arm. otx
“pied", comme tasn “ dix " et les themes en -i- et en
C'est que cette nasale se presente elfectivement avec I'articulation
dentale en celtique (gaul. netucton, etc. . . .), en germanique
(got. paw-rt, Ivan, etc. . . .), en lituanien, en slave {sun- devant
voyelle a cote de su- devant consonne. et autres e.xx.), en grec
(toV, etc. . . .), en armenien (cf. oln, tasn centre skr. pdilum, lat.
decern, etc. . . .), toutes langues auxqueUes s’ajoute maintenant le
hittite {mahlan acc., etc. . . .). Au contraire, elle n'affecte I'articulation
labiale qu'en indo-iranien et en italique. De plus, M. A. Meillet faisait
observer que chaque langue indo-europeenne prise a part ne presente
jamais qu'un des deux traitements, le traitement labial -m on le
traitemeut dental -n. Cette secoude remarque n'est contredite qu'en
apparence par les substaiitifs neutres en -en, -men du latin (ici
I'analogie a fait disparaitre les effets de revolution phonetique reguliere,
qui seraient *-cm. *-m.em, les gen.-dat. sing. : -inis, -ini : -minis, -mini
et tons les cas du pluriel. c.-a.-d. -ina. -mina, -inum, -minum, etc., dans
lesquels -n- etait interieur, ayant fait retablir -n au lieu du -m attendu,
et atteste par rombrien) et par les finales sanskrites du t\-pe de tnsmin,
locat. masc. -neutre et vdjin. vocat. masculin. etc. qui sont. evidemment,
d'origine receiite. cf. p. ex. avest. calimi, knhmi. (Aussi le samdhi
n’est il pas le meme ; tnsminn-eva mais tam-era.^)
E.st-ce a dire que 1' indo-europeen ne connaissait absolument que
-n. final et que, dans tous les cas de mots finissant sur une nasale, cet
-n ait ete mecaniquement transforme en -m, en indo-iranien (et de
meme en italique conimun on du moins en prelatin, cf. arch, ddnom,
^ Lt‘ Ninakrit connait des fi finaux d’origine recento {pmn. etc., neutre prak).
L’indo-europeen n’avait certaiiiement qiic -n ou -m en hii de mot. A plu< forte raison
n'avait-il en cette position ni n. ni n.
478
A. CUNY —
en preosco-ombrien, cf. o. Unvlanum “ Nolanorum ”, etc. . . .) ? Non,
car il est certain que I’indo-europeen disposait, non seulement de
themes, mais de formes nominales on verbales finissant les unes sur
-n, les autres sur -yn. Ce qui reste \Tai, c'est que chaque idiome,
pris en particulier, ne connait plus que -yn ou -n final. II faut au reste
distinguer ici entre suffixes et desinences. La oii il s’agit de themes
appartenant au genre inanime (neutre), le morpheme suflfixe finissant
siK une nasale : -n ou -m (soit -»» au degre zero) se trouvait etre
en finale absolue, et la chose se produisait egalement dans quelques
noms de nombre et dans quelques indecfinables, v. ci-dessous.
1. Desinences (il faut en outre distinguer ici entre desinences
nominales et desinences verbales).
A. Noyyi. — Parmi les differentes desinences nominales la premiere
a considerer est celle de I’accusatif singulier des themes masculins
et feminins en -a-, -a-, -o-, etc. (cf. aussi les themes m.-f. con-
sonantiques) et la nasale du nominatif-accusatif singulier des themes
neutres en -o- (finale sans doute inorganique). Il est bien certain que
c'est M. A. Meillet qui a raison et que c'est sous la forme Hon, *tdn,
*ag,royi, *owin, *bMg,him, etc. . . ., *ddnon, etc. . . ., qu’il convient de
restituer les prototypes des formes sanskrites tarn, tdyn, djram, dmin,
bdhum, etc. . . ., ddyiayyi n., etc. . . ., cf. gr. toV, r-gv (dor. rdu), lat.
{is)tuni, {is)tam, dypov, agruryt, ep. oiv, att. otv, lat. ouem, nrjxw, cf.
senatum, nmnuyn, etc., etc. . . .
Ensuite, il faut faire etat de la dfeinence de genitif pluriel : indo-
europeen -ow, gr. -wv, -6 jv, lat. -uyn, etc. . . ., soit un indo-iranien
tres antique *-a« et, avec repetition de I’indice de cas (mais on a encore
vM. devdyn et un ou deux autres exemples), la desinence ordinaire
-dndyyi des themes en -o- et en -a-, teUe qu’elle a ete expliquee par
M. A. Meillet (article cite). C'est done au cours de I'epoque indo-
iranienne qu'a eu lieu la transformation phonetique de -dn final en -dryi,
et le cas est le meme que celui de I’accus. sg. dont les formes telles que
uyiddm (accompagne d'un auxiliaire et suppl&nt le parfait de unatti
[Atharvaveda] ) n’est sans doute qu'une attestation particuliere : accus.
*uyidd-yi ; si cette forme existait deja en indo-europeen, elle a entraine,
en latin, la creation d'lm paradigme : uyida, undue, etc. . . . sur le
modele de aqua, aquae, etc.
Mais il y a aussi, dans les pronoms a genres et dans les pronoms
personnels, une sorte de desinence indo-iranienne -ayn (on la retrouve
egalement en latin dans id-em, de meme qu’en ombrien) ; on a : skr.
demonstr. nom. masc. ay-dyyi, fern, iy-dm, n. id-dm (cf. lat. is, ea, id),
LES NASALES EN FIN DE MOT EN SANSKRIT
479
pronom pers. ahdm, v.p. adam, avest. azsm “ ego ”, 2® pers. t(u)v-dm
“ tu ”, plur. p. vay-dm “ nous ”, yuy-dm “ vous ”, indo-iran.
*yuz-am de *yus + am, c.-a-d. *yus-9,en, v. ci-dessous — I’explication
est celle de M. A. Meillet qui, dans skr. yuydm, voit une assimilation
(au lieu du *yurdm attendu), etc. . . . On a meme le reflechi svay-dm,
mais la forme est indeclinable. Bien que -am se rencontre dans un
neutre (iddm, lat. id-em) et dans des formes de pluriel, on pent y voir
un indice de nominatif {mihi et tihl, etc., montrent qu’il est
analogique dans skr. mdhyam et tubhyam, on a aussi tubhya). C’est
I’equivalent semantique a la fois du -s du nomin. masc.-fem. sing., du
-es du nominatif plur. de meme genre et du -n du nominatif (acc.)
neutre des themes en -o-. La comparaison de cet element avec le ’an-
du semit. comm, dans *’an-’a “ego” (ar. class, etc., ’ana), *’an-ta, fem.
*’an-tt “ tu ” (ar. class, etc., ’anta, fem. ’anti), *’an-tun “ uos ” (ar.
class, etc., ’antum) et autres formes pronominales dans lesquelles le
morpheme ’an- occupe la premiere place au lieu d’apparaitre a la
seconde (fait assez frequent pour tons les morphemes en semitique)
prouve que, dans la restitution sous forme indo-europeenne de cet
element il faut se decider pour -a,ew. On notera que -n des neutres en
- 0 - pourrait tr^s bien representer le degre z4ro de cet element, mais la
chose supposerait I’amuissement de a,.
Le cas est done le meme que le precedent : I’indo-iranien et
I’italique ont transforme phonetiquement -9^en en -am (resp‘ -em). Les
autres langues ne semblent pas avoir garde de traces de cet -Sfin. [il
fonctionne comme pluriel aussi bien que comme singuher en indo-
europeen et en semitique. Sur ce dernier domaine il est meme admis
dans le duel ; ar. cl. ’antumd, mais le fait se retrouve encore en indo-
iranien : ved. dvdm, accus. dram “ nous deux L'identification
proposee est done des plus plausibles.]
B. Verbe. — Ici U faut d’abord retenir les desinences de duel aux
2® et 3® personnes secondaires de I'actif : skr. -tarn, -tdm (on
en a I’equivalent en \deux-perse et en avestique, cf. aussi gr. -tov, -rdv
(Pindare), att. -ryv. Le vieux-slave -ta (de meme le lituanien) ne presente
pas de nasale finale, non plus que I'ombrien -tu-ta (si Ton adopte les
vues de M. Nacinovich, Carmen Arvale, qui y voit d’anciennes desinences
de duel, cf. lat. -tis pluriel, mais skr. -thah duel) et ne decide natureUe-
ment rien au sujet de la qualite de la nasale finale. Mais il y a toute
chance pour qu id nous ayons affaire a d'anciens *-to-n, *-td-n indo-
europeens, etant donne le hittite -ten (devenu pluriel dans I’usage,
p. ex. iyatten “ vous faites ”, cf. aussi le semit. *-tun (hebr. -ten, etc. . . .)
480
A. cuxy
et peut-etre nieme le hittite hierogh-phique -tana dans rastana d'apres
B. Hrozny, Les Inscriptions Hittites Hieroglyphiques, P Lieferung,
p. 93 (niais M. P. Meriggi ' n'est pas de I'avis de M. Hrozny). A plus
forte raison songera-t-on a *-tan pour la 3*^ sing, de I'imperatif medio-
passif (skr. kriya-tant) qui n'a meme peut-etre pas une anti quite indo-
iranienne, et a *-d}mon pour la dfeinence secondaire de deuxieine
personne plur. au nioyen, desinence indo-iranienne que Ton retrouve
presque identiquement en hittite et qui peut-etre etait duelle autrefois,
cf. gr. -aBov. Comme ici -n etait inorganique. cf. 2® primaire plur.
inoyen skr. -dhre, il n'y a pas de raison pour y voir autre chose que -n
a I’origine, cf. les neutres en -o-.
II en est en revanche tout autrement de la desinence secondaire de
premiere personne singulier actif. En effet. .soit au present (lat. sum,
o. sum, indo-europ. *s-om, existant a cote de *es-»u suivant les vmes
de JM. J. Bonfante, admises maintenant en partie par M. A. Meillet
dans son Introduction, ed.), soit a I’imparfait, qu'il s’agisse d’un
thematique : skr. dbharam, gr. €<f>epov, etc. . . ., ou d’un atheinatique
(.skr. dcrunavani; gr. ihelKvvv, etc soit dans les divers aoristes,
thematiques, p. ex. dvoccim (analogique pour *dvokam, cf. av'est.
vaocat) ou athematiques {drautsam, ddarqnm, etc. ... — pour dqrunavam,
drnutsum. etc., cf. dans les noms pad-am comme dgva-m. mais gr.
TToSa et L-n-TTov — , il parait certain que la finale primitive etait -m,
-m. soit done -o-m dans le t^’pe thematique, que Ton pense. comme
hauteur de ces lignes. ou que I’on nic, ainsi qu'on le fait generalement,
qu'il y ait un rapport etraiologique entre cet -/« {-in) et le *we qui
est a la base des formes du pronom singulier de I"'-' personne ; skr.
gen.-dat. atone me, gr. poi, hit. wu, skr. dat. tonique mdhyam, lat.
mihl ; .skr. accus. tonique mam, v. si. me, lat. me, gr. qe, etc. . . .
2. Morphemes noniinaux suffixes (cas de la desinence zero)
iSi les mot.s qui coinportent des .suffixes a nasale -n, -m appar-
tiennent a un des genres animfe, p. e.x. *yjiyem- ” hiver " (fem.),
*(jffiem-, *yjieni- terre ’’ (fem.). *sem- arinec, etc " (skr. sdm-d,
etc fem.). ils nc viennent pas ici en consideration, car la nasale qui
termine le theme ne sc trouve jamais j)lactA' on finale ab.solue. On pent
signaler sculcmcnt que, pour le premier de ces mots, *yjiyem-, iM. jM.
Ernout et Meillet (v. leur Dictionnaire. p. 431) admettent I'exi.stcnce
d un second theme (a iia.sale dentale), soit *yjiyen- et que dans
Archil Jur Oricnlfurschung, t. x, 3® cahier, Berlin, 193-5, p. 121.
LES NASALES EN FIX DE MOT EX SANSKRIT
481
Emerita (Madrid, n° de decembre 1935), il est montre que le benefice
de cette observation doit etre egalement etendu aux noins de la
“ terre ” et de V “ annee ”, soit done des formes paralleles *(j\h‘<on-
(gr. x^ova, etc. . . .), *sen- (lat. senium, senecta, etc. . . .). Car le fait a
de I'importance pour le nom de nombre “ un ” (en grec efb', jxla, ev).
D'apres I'explication courante, il s’agirait d'anciens *sem-s. *sm-i>j:,
*se-m. Mais. plus probablement, on av'ait d’abord *scn-s masculin et
surtout *sen neutre (nominatif-accusatif). Si on I'admet, on sera
dispense de recourir au detour propose par M. A. ileillet (art. cite,
p. 372) : ” Le passage de -m a -n doit etre tenu pour anterieur a
I'existence separee des dialcctes hi.storiquement connus.” En effet,
dans cette hypotliese. *sem (nomin.-accus. neutre sing.) aurait passe
a *sen dans la langue conunune, pour redevenir *$etn en italique et
*sam en indo-iranien (cf. skr. preverbe sum- " av^ec ”, litterafi “ en
un”, lat. sem-el. arch, semul, class, simul, etc.). Au reste, *sen originaire
etait deja admis par R. Gauthiot (Fin de ynot, 1913) et par M. J.
Mansion, Museon, 1913, p. 250). Suivant done qu'on partira — pour
rendre compte du skr. sam, — de indo-europ *sent. ou *sen, qui
sans doute e.xistaient cote a cote, on constatera une conservation
ou une transformation phonetique de I'etat indo-europeen. Mais,
pour les derives ; skr. samd-, etc. . . ., lat. similis, etc., il faut surement
partir de *sem coniine base,
Dans le cas des substantifs de genre neutre (ceci serait vrai du reste
du nom.-acc. neutre d’adjectifs du ti-pe correspondant au nom. masc.
sing, sdmardjd p. ex.) : skr. ndma, <;drina. *gr6mu (on n'a que
gromatayn, cf. v. h. all. hliuyytunt “ renommec), etc., gr. dcoga, Aetp/ia,
etc. . . ., lat. iyiguen (serait en grec *d/3a, cf. le masculin regulier
d§7)v < *ng",eyi). lat. laiguen (cf. yuigueyytuyyi. skr. dnjaii, v. b. all.
ayicho masc.. etc.), il est absolument sur que I on doit partir de themes
finissant par la nasale dentale, .soit done *leik‘[-ynn.
*iig'‘jj, *oiig“,>y (*oiig“yiton), etc., etc. On a naturellement ici -a comme
finale en Sanskrit. [I! est a noter que si I'analogie avait amene la
creation de *-an (cf. le cas de pddum. etc drautsany, etc. . . .), cet
*-iin serait phonetiquement devenu *-a//(.] De la sorte yuiyna et autres
neutres analogues se comportent exactemeiit comme yidvn " 9 ' (*neicyi)
et comme saptd " 7 " (*septyn) et ddsa - 10 " (*dek\yu). En latin, .seuls
*ndyy>eyyi. *ingueyn. *uyigui'yn. etc., auraient ete reguliers, mais on a vu
que I’analogie avait fait di>paraitre ces finales, rombrieii gardant
toutefois des traces du traitement phonetique regulier ; malheureuse-
meut dans cette langue, comme dans le cas des accusatifs sg. masc.
482
A. CUNY—
et fern, et dans le cas des nom.-acc. neutres en -o-, le -m final est tres
sou vent omis dans la graphie, mais on a numem “nomen” bien
atteste sur les tables en ecriture etrusque.
Noms de nmnbre et mots isoles
Ces mots etant indeclinables, il ne pouvait etre question d’influences
analog! ques provenant de la flexion. Ains i *septm, *dek,m et *nemn
sont done indifferemment saptd, ddsa, ndva en Sanskrit, septem>
decern, nouem en latin (ceci malgre lat. non-us, got. niun, lituan.
devintas “ neuvieme ”, opposer btuan. desimtas “ disdeme ”), tout comme
gr. eVra, Seica, (ev)i>ea, arm. eu’thn, tasn. A noter que la plupart des
faits r&umes jusqu’iei sont bien connus, mais il faUait les remettre
en memoire pour bien montrer que, dans la question du traitement
des finales, il ne faut tenir compte ni des formes nominales du genre
de pddam, etc., ni des formes verbales du genre de drautsam, mais
envisager uniquement les finales (indo-europeennes) -in, -un, -bn, -an,
-9ji, -en, e’est-a-dire indo-iraniennes ; -hn, -um, -am, -dm, etc. (le
Sanskrit comme le latin n’a conserve que les themes en - 6 -, tandis que
le grec a encore les deux, en -o- : naXio-s “ cable ” en attique et en -o-
KaXos “ cable ” en ionien, etc. . . . ; de meme -»n (gr. -dv) ne parait
pas avoir de repondant en Sanskrit, opposer vrkim a Xvaaiv).
Pour les mots isoles, on pent rappeler *kon,^ skr. kdm, v. si. ku
(et kun devant voyelle, cf. p. ex. russe kn-jemu “ a lui ”), lat. cum
(con- et CO- en composition), *k"on (lat. quom, cum, got. han), *ton
(got. pan), lat. turn comme num (cf. tun-c et nun-c) soit *tu-n et *nu-n
(gr. I'w, cf. skr. tu, nu) ; *k'"-dn (lat. quam, d'oii quan-de, quan-dd,
armen. khan). Le cas est le meme que pour toutes les categories
precedentes (sauf celle de *[e\bherom, skr. dbharam, gr. ecfiepov et autres
formes analogues). On a -n -* -m en indo-iranien comme en itaUque.
II. Articulation indo-iranienne des nasales en fin de mot
Le traitement indo-iranien et le traitement italique (coimu surtout
par le latin), bien qu’il y ait entre eux de grandes ressemblances, ne
sont pourtant pas de tout point identiques. C’est quand il s'agit
de la phonetique de la phrase (sarpdhi = phonetique syntactique)
qu apparaissent les differences. Une citation quelconque suifit a
le rappeler, soit par exemple le sloka emprunte par A. Bergaigne
(Manuel, p. 2) aux Indische Spriiche de Weber :
II oil gr. Koivos < *kon-yus. Car ^alvcu represente *g'in.y6, ainsi que le montre
le lat. uenio, ueni, osque klim-bened.
LES NASALES EN FIN DE MOT EN SANSKRIT
483
tenadhitam srutam tena term sarvam-anusthitam j
yenasdrn prsthatah krtvd nairdsyam-avalarnbitam |i
a cote de ce vers de I’Odyssee (a, 1) :
dv8pa yoL ewene, Movaa, noXvrpoTrov, og p.aXa ttoXXo.
et de ce vers de Virgile {Eglogues, x, 13) : —
etiam laur' etiarn Jleuere rnyricae.
Dans le texte sanskxit sarvam devant anusthitarn et nairdsyam
devant avalamMtam sont traites de la meme fa 9 on exactement que
TToXvrpoTTov dans le texte grec devant o? (mot a initiale purement
vocalique, si Ton tient compte de la psilose eolienne ou ionienne
masquee par la graphie attique avec esprit rude, h = ■). Dans le texte
latin au contraire, la finale devant etiam est traitee exactement
de la meme fa§on que la finale ‘ de Zowr* devant c’est-a-dire
qu’elle est “ escamotee ” dans la prononciation (fausse elision, suivant
I’enseignement de Louis Havet, mais reelle disparition).
Pour le vieux-perse et I’avestique la graphie ne nous revele pas
qu'une finale -am (par exemple) soit traitee differemment selon
qu'une voyelle ou une consonne ouvre le mot suivant dans la phrase.
Ainsi adam ddrayavaus est note de la meme fa 9 on que s'il etait suivi
d’un mot a initiale vocalique, p. ex. avam asmatiam, Meillet-Benveniste,
Gr., p. 11, comme adam Bardiya a{h)miy, ibid., p. 135 (de meme pour
avest. azam).
Outre I’option pour la nuance labiale -m devant voyelle en sanskxit
(et partout en latin), ce qu’il y a de commun entre le traitement Sanskrit
de -arn (par exemple) et le traitement latin de -urn (par exemple)
en cas d’initiale consonantique du mot suivant — le choix de la nuance
labiale, -arn, est uniquement le fait des transcripteurs europeens,
mais ils paraissent avoir ete guides par un sens exact de la langue —
le trait commun de ces traitements, c'est I'affaiblissement (il se
manifeste, on le sait, dans toutes les tranches finales de mot, v.
Gauthiot, Fin de mot). Get afifaiblissement a entraine les consequences
que void : en sanskxit comme en latin le complexe phonetique forme
par voyelle + nasale finale s' est reduit a une simple voyelle nasahsee
du genre des voyelles fran 9 aises -an, -in, -on ou des voyelles polonaises
f, p, mieux encore de -an, -in, -on fran 9 ais articulfe par des sujets
parlants originaires du Sud-Est de la Prance, p. ex., pc'" au lieu de
jre (ecrit pain). Toutefois I'affaiblissement etait bien plus considerable
en latin. Les grammairiens indigenes signalent le fait et tres souvent
-m final n'est pas note sur les anciennes inscriptions (meme fait en
484
A. CXnSTT —
oinbrien, p. ex. ivengaru = lat. iuuencanim, etc. . . .). Malgre des
juxtaposes tels que quamohrem, quemadmodum, pour lesquels les
Anciens ne nous disent pas si I on escaniotait -am, -em ou si I on
articulait qua-mo-hre{)ik), que-mn-dmo-dii(m), la po&ie classique prouve
qu’une voyelle quelconque suivie de -m final s'elidait (ou niieux
s'escamotait), meme en cas de monosyllabes : ici on evitait generale-
ment I'elision, rem par exeniple etant reduit a une seule consonne r-,
et cela comme si le mot finissait directenient sur une voyelle. En un
mot, I'affaiblissement des finales latines -um, -im, -em, -am, etc.,
etait aussi considerable que celui des finales arabes -u’', -i”-, -a" nieme
en ar. classique (cf. vieux-babylonien -tim, -im, -am, sabeen -um, -im,
-am, semit. comm. -*u'^ 'm, *-i” m, *-a" in). Ici I’affaiblissement etait
si sensible que, meme quand il s’agit de textes anciens, les arabisants
traitent aujourd’hui ces finales comme n'existant pas, qui pis est,
comme n'ayant jamais existe, ce qui est aussi contraire a la realite
liistorique que si Ton enseignait qu'en latin -um, -im, -am, etc., n'ont
jamais ete prononces, parce qu'en fran 9 ais ils sont representes par
zero ou par e muet et qu’ils ne le sont que par de simples voyelles -o, -e,
-a en espagnol, italien, etc. . . ,
Vu cet affaiblissement, la veritable notation pour le latin serait
u, 1, 0 , a, e, soit de simples voyelles nasalisees. Ceci veut dire que le
stade intermediaire (-u”') entre -um, etc., et u, etc. . . . etait deja
definitivement depasse (au moins a I'epoque classique : i^'*' siecle avant
notre ere). Au contraire, le .Sanskrit (vMique. etc. . . .) en etait encore
exactement au point moyen de revolution. En effet, on voit tres bien,
si on I'admet. pourquoi. devant initiale vocalique de mot suivant. on a
en Sanskrit -um, -im, -am. etc. . . . C'cst qu’il s'agit ici d'anciens -u'",
-d"‘. etc. (cf. les transcriptions -)/w. -ini. -nni, etc., devant consonne
alors que peut-etre il ne s'agit plus deja que de -ii. -1. -d. etc. . . . tout
ceci malgre la prononciation de -m en -n reconnue niaintenant, mais
pour le tokharien seulement). Dans ces finales, -u™, -T"‘. -d'", le "
a ete reiiforce par 1 attaque vocalique (douce) du mot qui suivait.
Cet une fois consolide (au contraire devant initiale con.sonantique
il .s'etait sans doute evanoui i) a denasalise les il, I, a, etc., qui le
precedaient immediatemeiit [plienomenedef/i^eVe/icir/ffoa, v. A. iMeillet,
M.S.L.. t. xii. pp. 14-34], d'oii. en definitive, dans la phrase,
-um, -im. -am. etc,, devant voyelle. En grec, revolution a du
etre la meme. a part ce detail Cjue la langue avait, au rebours
de 1 indo-iranien. opte pour I'articulation dentale de toute nasale en
^ Xc* lai'^ant bubsister que la nasalisation.
LES NASALES EN FIN 1)E MOT EN SANSKRIT
485
fin de mot et que, comme le vieux-perse et I'avestique, le grec ne fait
aucune difference entre le cas d'initiale vocalique et celui d'initiale
consonantique du mot suivant, employant partout la script io plena
(et sans doute aussi la prononciation inalteree de la nasale en fin de
mot comme a I'interieur).
En grec on a done eu, comme en sanski-it, devant voyelle, un stade
-0", -d”, -S", etc. . . puis -on (-ov). -an {-av), -un {-vv), etc. . . . de-
nasalises par le meme precede de differenciation que ci-dessus, cf.
p. ex. TToXvTpoTTov dcvunt os, soit [^]os. Toutefois — on l a deja
rappele — ce traitement, apres consolidation de en -v, a ete etendu
en grec a tons les cas. De meme, I'aspect sourd de la " sistante $
(terminologie de F. de Saussure), soit -s, aspect qui n'etait legitime que
devant occlusive sourde, a ete generalise (il n'y a des traces de I'aspect
sonore ancien, -z, (sous forme de -p) qu'ici dans quelques inscriptions
dialectales : laconien, eretrien, cleen, tandis que le Sanskrit, on le salt,
distingue toujours entre les cas d'initiale sonore et initials sourde du
mot suivant, p. ex. -ah {-as) et -o, de meme que dans le cas de -m, il
distingue toujours entre initials vocalique et initials consonantique.
Devant initials consonantique, I’accord du Sanskrit et du latin est
presque parfait : etani (vlrani) p. ex. comme ilium (uinim) p. ex., avec
une voyelle nasalises plus ou moins affaiblie, le latin toutefois etant
beaucoup plus evolue. Devant voyelle, la finale nasalisee a continue
de s'affaiblir en latin et a fini par ne plus compter, tandis que, par
la voie indiquee ci-dessus, elle a ete en fin de compte, retablie en
Sanskrit.
Ce qui est done indo-iranien dans le traitement indien des nasales
finales, e'est I'option pour -m seul (au lieu de -m et -« primitivement
existants). Ce qui appartient en propre au Sanskrit, e'est I'emploi
de -am devant initials vocalique (mais aussi en fin de phrase) et de
-am devant initiale consonantique, alors que I'iranien ne semble faire
ici aucune distinction. Cette distinction est-elle ancienne ou non ?
K. Brugmann. dans son Grundriss, enseignait que les traits essentiels
du samdhi (vedique et classique) remontaient a I'indo-iranien et
meme a I'indoeuropeen. Il avait sans doute raison, ainsi qu'on le
voit par la chute indoeuropeenne de -m, -n, -r en fin de mot apres
voyelle longue ; vedique hsa-h " terre". c.-a.-d. preved. *l:sd (avec -s
surajoute), issu d’un nomin. indo-europ. *g,:hd (pour la voyelle, cf.
xdd>y). avec chute de I'-w de otn- apres allongement de la vovelle
thematique ; de meme avest. zgd "hiver”. ancien *jhgd (avec -s
du nominatif surajoute) issu d'un nominat. indo-europ. *gjujd (pour
486
LES NASALES EN FIN DE MOT EN SANSKRIT
la voyelle, cf. chute de I’-m de ^gjiy^^jotn- apres allonge-
ment de la voyelle thematique), puis tout le type de skr. adhva
“ chemin ”, theme ddhvan-, v.-lat. Jiemo, lat. homo, sermo, etc. . .
anciens nominatifs en -6 avec chute indo-europeenne de -n apres
allongement de la voyelle thematique, cf. encore pita, de indo-europ.
*pHe, avec chute de -r final apres allongement de la voyelle du theme
au nomin. singulier. Si done -m, -n, -r tombaient en fin de mot apres
voyelle longue (*g,zhd < *g,zhd’^, etc.), e'est la preuve que -m, -n, -r
etaient faibles meme apres voyelle breve. Au reste, I'indo-europeen,
dans les cas cites plus haut, parait avoir generalise la forme a nasale
(ou a liquide) completement reduite. L’etat indo-europeen des nasales
en fin de mot devait done etre a peu pres ce qui suit (par convention
la voyelle o representera ici une voyelle quelconque) :
1) devant initiale consonantique -o" et -o'".
2) devant initiale voealique -on et -om}
A I’epoque indo-iranienne on eut :
1) -o'" (avec identification des deux finales, devant consonne).
2) -am (avec identification des deux finales, devant voyelle).
[En iranien le traitement 2) parait avoir ete generalise comme
en grec.]
De meme a I’epoque italique (commune) on eut :
1) -o'" (d'ou : lat. arch, -mn et -o, class, -um devant consonne).
2) -om (et lat. class, -um, note de meme, mais tres faible ainsi que
le prouve I'elision en poesie).
En resume, ainsi qu’il arrive souvent, le Sanskrit, a I’interieur de
I’indo-iranien, et le latin, a I’interieur du groupe italique, ont evolue
par des voies tres analogues et se montrent tres conservateurs, la
surtout oil ils sont d'accord. L'innovation, qui a introduit -m au lieu
de -n s’est produite, la chose va de soi, independamment en indo-
iranien et en italique. ^ Quant a I'affaiblissement du complexe voyelle
-r nasale finale, le latin qui I'a maintenu et peut-etre exagere, parait
avoir ete plus conservateur que le Sanskrit, qui Test a son tour plus
que I'iranien (et que le grec).
^ Dans beaucoup de parlors » et w consonnes, comme elles le sont ici, nasalisent
la voyelle precedente. Ainsi dans tout Je fran^ais de I’Est (au centre aussi :
Pithiviers). on prononce p. e\. il ?nhi(p) au lieu de il etc., etc.
" Cot le soul point .sur laquelle la tradition a ete alteree en latin et le fait n’est
pas tres ancien ainsi que Tindique lat. quon-iam a cote de quom, cum. Le Sanskrit
a sans doute cede aussi sur celui des voyelles nasalisees bnales devant vovelle. —
J aurais du citer plus tot La nasaliie en indo-aryen de M. J. Bloch, article paru
dans le volume du (J \r,qua7itenaire de VErole des Etudes (1918).
Der Typus tuda- im Altindischen
Yon Albert Debrunner
D ie folgende Untersuchung ist aus der Arbeit an Band ii, 2, von
J. Wackernagels Altindischer Gramtnatik hervorgewachsen,
der die nominale Ableitung behandelt. Das Material ist gewonnen
aus dem Kontrarindex in Grassmanns Worterhuch zum Rigveda,
aus AYhitnejs ,, Index verborum zum AV.“ (JAOiS.,12), aus Whitneys
,, Boots “ (Anhang ii zu seiner Grammatilc), aus den einschlagigen
Regeln des Panini (nebst Mahabhasya, Kasika und Siddhanta-
Kaumudl) und aus Wackernagels wertvollen Sammlungeii zu diesem
Kapitel. Jeder, der selber auf dem Gebiet der ai. Grammatik arbeitet,
weiss, dass ein so gewonnenes Material Liicken aufweist ; sie waren
nicht wesentlich kleiner geworden, wenn ich alle Stichworter der
Petersburger AVorterbiicher durchgesehen hatte. Das Gesamtbild
diirfte trotz der Liicken in den Grundziigen richtig sein. Fiir die
Verarbeitung und Darstellung trage ich die Verantwortung allein.
Ich gebe zuerst eine alphabetische Liste aller irgendwie in Betracht
kommenden Wbrter, Sicheres, Zweifelhaftes und Falsches neben-
einander. Dabei sind die M'orter, die sicher oder sehr wahrscheinlich
nicht zum Tj-pus tudd- gehoren, in Kleinsatz gedruckt ; diese AVorter
sind in der nachherigen Behandlung nicht beriicksichtigt.
Mit Ai. Gr. wird im Folgenden auf die erschienenen Biinde
von AVackernagels AYerk verwiesen ; nach diesem sind auch die
Umschrift und die Abkiirzungen gestaltet.
Aiphabetische Liste
ep. ihga- ,, beweglich “ ; v, ing- ,, sich regen “ (Pras. ep. kl.
ihga-).
V. ind- ,, stark, tiichtig “ : in- in SB. upenita- ,, eingedriickt "
(SA". 1, 2, 2, 4, 2 = 1, 176a inimasi fiir RA^. 10, 134, la minlmasi ?) ;
vgl. aw. a{i)nita- aus *an-inita- ,, nicht gekrankt “, in-ti-
,, Krankung “ ; aber sonst v. inoti, invati ,, treibt “ und dazu -invd-
(AA'ackernagel, Ai. Gr., ii, 1,181).
V. gav-isd- ,, Kiihe begehrend “ (= v. gav-is-), v. presa- ,, Drang “
( = V. pres-) : v. is- ,, 'wdinschen “ (Pras. isa- erst ep.).
-Iksa- Y. 8 zu P. 3, 2, 1, dazu Pat. sukha-pratiksa- ,, Gliick
erwartend “, andere A'orderglieder mit prafiksa- ep. kl. : nicht zu
V. iks- (Pras. tksa-) ,, sehen “, AV. prati Iks- ,, erwarten “, sondern
488
A. DEBRUNXER —
Bahuvrlhi-Hintergliedform von U. ep. pratiksd- ,, Erwartung “ ; SB.
tksa- ist nur etyniologische Spekulation.
-Ihkkd-: V. prehkM- ,, schwankend, Schaukel “ ; v. pra liikh-
(Pras. AA. tiikha-) ,, schiitteln, schaukeln “ (vgl. v. -inkhayd- aus
dem Priisensstamm Ai. Gr., ii, 1, 180).
ep. Ira- ,, Wind “ : v. ir- (Priis. tra-) ,, in Bewegung setzen “.
Vgl. aw. Ira-, n. ,, Anlauf, Tatkraft
SB. Iso- „ Herrsclier ■■ (= VS. U-), U. als Hinterglied : v. is-
„ herr-scheii ” (Priis. v. eininal isate neben gewohnlichem Iste).
kl. Via- „ Verlangen " : B. Ih- (Priis. tha-) ,, verlangen “.
B.S. -uksa- ,, besprengend “ (= v. -uks-) : v. uks- (Priis. uksd-)
„ besprengen “. Anders hrhdd-ukm-, s. Ai. Gr., ii, 1, 116, und Keith
zu TS. ], 4, 266.
S. -njjJia- „ aufgebend " : ep. kl. v.jjh- (Pras. ujjhd-) „ aufgeben “.
S. uiicM- (Oxytonese nach P. 6, 1, 160) ,, Nachlese “ : S. kl.
uiich- (Pras. iihcha- und unchd-) „ auflesen “.
-iida- in P. Vop. avoda- : v. iid- ,, benetzen “ (P. 6, 4, 29 aus v.
und- mit Nasalschwund) ; dock eher Bahuvrihi mit -ud-a- (Ai. Gr.,
ii, 1, 92).
-ubjd- : V. uruhjd- ,, weit geoffnet “ aus *ud-ubjd- (Ai. Gr., ii,
1, 12), S. ny-ubja- ,, nach unten gekehrt “ ; v. uhj- (Pras. ubjd-)
,, niederdriicken “.
upi- ., leuchtend BR. ist Irrtum (usena RV. 2, 2, 86, statt
arusena) ; Lex. upt- ,, Tagesanbruch Susr. usd- ,, Brennen “ ;
V. ros- os- ., leuchten •' ; iiber u.pi- ,. Morgenrote “ s. Ai. Gr., iii, 283.
iisa- ,, bcgierig R\ . 10, 9o, 4a (?), ,, Liebhaber Lex. : v. loi- us- „ begehren “ ?
(nach Grassrnann Wb. s fur $ nach Ai. Or.^ i, 226).
(1) -uhd- in AV. sam-uhd- upokd- „ Anhiiufung “ : v. uh- (Pras.
uha-) ,, schiebeu
(2) ep. iiha- ,, tlherlegung M. kl. dur-uka- ,, schwer zu
begreifen " ; v. fth- (Pras. v. ohnle, ep. kl. ukate nhati) ,,uberlegen “.
fk-m- ,, ^■e^de^be^ " (?) RV. 8, 24, 27a : AV. rots- ,, verderben “ (?).
-fdha- ; sam-fdha- ,, zusammengefiigt *' (zu rdh- ,, gelingen, fordern ‘‘) wird aus
xamfdhf, ra R\'. 7, 103, oo ersohlos.sen ; richtig ist aber Ann.ahme von Instr.
samfdhd (mm-fdh- ,, gutes Gelingen " RV. 6, 2, lOe), s. Oldenberg z. ,St.
Lex. kird- (auch kiri- und kiti-) „ Wildsehwcin “ nach P. 3, 1, 13d aus v. kf-
,, aiis.strouen “. Sehr fraglieh.
ep. kura- ,, weibliche Bru.st “ : S. kuc- (Pras. kited-) ,, sich zusanimenziehen “ ?
V. t«/so- Eigenname : ep. kl. kutsdya - ,, schmahen ?
B.S. kupii- ,, Wagelialken '' : v. kup- ,, in Aufregung geraten ‘L
ep. kl. kuja- „ Gemurmel " : AV. kuj- (Priis. kuja-) ,, knurren “.
S. kurdu- ,. Sprung ; ep. kl. kurd- (Pras. kiii'da-) ,, aufspringen “.
DER TYPES TUDA- IM ALTINDISCHEN
489
V. 'krsd- ,, mager “ (=jungaw. kgrasa-, vgl. tscliech. krs
„ Zwergbaum “ u. dgl. bei Walde-Pokorny, Vergl. Worterb., i, 420) :
V. krs- ,, abmagern
•kra- : RV. 1, 120, 2c dkrau ganz dunkei, gewohnlich als d-kr-a- ,, untatig “ ( : v.
hr- ,, machen “) erklart.
V. kridd- ,, spielend YS. pra-krldd- „ Spiel S. ,, Spielplatz “ :
V. krld- (Pras. krtda-) ,, spielen
ksipa- ,, werfend “ Siddh.-K. Nr. 2897, vi-ksipa- Kal zu P. 3, 1,
135 : V. ksip- (Pras. ksipa-) „ werfen
-khidd- in a-khidd- „ an sich reissend “ MS. 2, 9, 8 (127, 3) — Kath.
17, 15 (258, 11), pra-khidd- ,, verzehrend “ MS. ebenda. dafiir Kath.
vi-khidd- ,, zerreissend (Mantra's) : v. khid- (Pras. khidd-). Alter
ist V. -khadd- aus der Vollstufe v. khad-.
-gird- -gild- : sam-gird- AV. 6, 135, 36 (vgl. 3a sdm girdmi !)
(verdorben, 18, 4, 606, smigirdh fur sanigiram des RV.), a-sani-sukta-
gild- ., Unzerkleinertes verschlingend ‘b 11, 2, 306, -gila- Y. 7 zu P. 6,
3, 70. gila- ,, Krokodil Siddh.-K. Nr. 2919 : v. gr- (Pras. AY. gird-
gild-). Ygl. auch V. -gir- ,. verschlingend “ und unten -gra-.
Lex. gunja- ,, Gesumme “ ; kl. gunj- (Pras. gudja-) ,, sutnmen “.
AV. gulphd- (v. 8B. kulphd-) „ Fussknochei “ ; S. vigulphaya- „ claranreihen “ ?
guha- ep. kl. als Name, kdka-guha- ,, Krahen verbergend " Pat.
zu V. 2 zu P. 3, 2, 5 : v. guh- (Pras. guhd- und gCiha-) ., verbergen".
Dock ist kdka-guha- eher Bahuvrihi ; „ den Krahen als Yersteck (v.
giih-) dienend “. Kl. -guha- \Yhitney, ,, Roots “.
grbhd- RY. 7, 21, 2c, wohl ,, ergreifend “ (vgl. v. gfbh- ,, Zugriff “),
dafiir 10, 119, 13a grhd- (s. Ai. Gr., i, 251) in der Bedeutung ,, Diener “
(oder auch hier = ,, Haus “ ?) : v. gr(b)h- ,, ergreifen “.
*grdhd- : v. grhd- m. ,, Haus “ = aw. garaba- ,, Hohle (als
Wohnung daevischer \Yesen) “ : idg. *gherdh- umhegen ". Mit
grhd- ist \’ielleicht gleichzusetzen Sanih. gelid-, n. ,, Haus (daraus
V. gehyd- ,, Hausrat '’) ; uber das auffallige e = r s. Ai. Gr.. i, 39.
-gma- : su-gtnd- zu v. gam- Benfey, Vollsl. Gramm., 135, § 368, 1 B 1 ; woher ?
-grd- ,, verschlingend “ in tuci-grd- ,, gewaltig verschlingend "
RY. 1, 140, 96 ; v. gr- ,, verschlingen ”. Ygl. tuvi-gr-i- 2, 21, 2c ;
oder ist dies = ,, laut schreiend (Geldner, Ubersetzung), also aus
V. gr- ,, s ingen “ ?
-ghusd- ,, tonend “ in aram-ghusd- „ laut tonend “ (?) AY. 10, 4,
4a : V. ghus- ,, tbnen “.
-ghnd- kommt in doppelter AYeise vor :
1) Als substantivisches Neutrum im Sinn eines Nomen actionis
VOL. vm. P.LKTS 2 AXD 3. 32
490
A. DEBRUNNER —
(vgl. At. Gr., ii, 1, 108 If.) 1, d. h. als Ableitiing mit a aus Komposita
auf -hdn- ,, totend “ : su-ghndya ,, zu leichtem Erschlagen “ RV. 8,
59 (70), lid, go-ghid- iind punisa-ghnd- ,,das Toten von Rindern,
Menschen “ 1, 114, lOcs, 'pamaya-ghnd- 10, 48, 8d, dhi-ghna- 6, 18,
14a, *h'a-ghnd- ,, die Kunst, den schlechten Wiirf (den , Hund ‘)
zu toten “ in v. svaghntn- ,,diese Kunst verstehend “ (E. Sittig
KZ. 52, 209). Vgl. aw. varad'ra-yna- ,, das Niederschlagen der Gegen-
wehr “ (Benveniste-Renou, Vrtra [Paris, 1934], 117) ; aber vrtraghne
RV. 5, 86, 3(Z ist nicht Lok. von -ghnd- (Hillebrandt, Ved. Myth.,
1, 191, A. 4), sondern Dativ von -han- (Oldenberg z. St.), ebenso
AB. 8, 23 (BR.).
2) Als Nomen agentis : d-puru-sa-gkna- RV. 1, 133, 6 f, hasta-
ghnd- ,, Handschutz (gegen das Aufschlagen der Bogensehne) “,
eigentlich ,, Handschlager “ ? 6, 75, 14c, *asva-ghnd- als Name vor-
ausgesetzt durcb den Namen dsvaghnd- 10, 61, 21d, pdni-ghnd-
,, Handeklatscher “ VS. 30, 20 (TB. 3, 4, 1, 15 in ahnlicber Umgebung
pdni-sam-ghdtd-, P. 3, 2, 55 und Lex. lehren pani-ghd-), Kaus. rahso-
ghna- „ Raksas totend S. ep. kl. vi-ghna- ,, Zerbrecher, Hemmnis
M. bhruna-ghna- ,, Toter der Leibesfrucht “, Bhag. kula-ghna- ,, das
Geschlecht vernichtend “ (Fern, -ghni- Ram., -ghnd- Mbh.), ep. tamo-
ghna- ,, Sonne “ (,, Finsternisvertreiber “). P. 3, 2, 52-4 ; 3, 3, 83 ;
3, 4, 73 lehrt die meist sonst nicht belegten Komposita mit jdyd-
„ die Gattin totend “ (Gegensatzbildung zu S. pnti-ghu- ,, den Gatten
totende Frau “, v. d-pati-ghnl-, wo natiirlich -ghni- das Fem. zu -hdn-
ist), mit andern Wortern, wenn der Tater kein Mensch ist, mit hastin-
,, Elefant “ und kapdta- ,, Tiirflugel “ im Sinn von ,, zu schlagen
vermogend “, mit stamba- ,, Grasbiischel “ im Sinn eines AVerkzeugs,
ferner upa-ghna- im Sinn von ,, Lehne, Angrenzung
kl.jira - ,, Kiimmel “ (auch jarana-^jima’ u.a.) : v. jr - ,, zerreiben “ ? (BR., Renou,
Gramm, sanscr., 214).
V. jtvd- ,, lebendig “ = aw. jva- (lies *pva-), ap. jiva-, lat. vivos,
lit. gyvas, abg. zivu, usw. : v. jlv - ,, leben “ (Pras. v. jiva-, aw. ](i)va-,
ap. pva-, lat. vivd, abg. zivo).
-jusa- : SB. aldm-jusa- ,, fiir sick ausreichend “ : v. jiip (Pras.
jusd-) „ geniessen “, vgl. auch v. -jup „ Gefallen findend “.
kl. jrmbha- ,, das Gahnen “, R. ,, ein Tier “ : v. jpnbh- (Pras.
jnnbha-) ,, gahnen “.
-jya- in AV. brakma-jyd- ,, Brahmanen unterdruckend “ nicht von v. ji- ,, uber-
waltigen sondem von der Torm jyd- derselben Wurzel, also wie v. -da~ von dd- usw.
-lira--. TA. pra-tird- „ fordernd “ : v. tf- , , hiniibergelangen,
DER TYPUS TUDA- IM ALTINDISCHEN
491
-bringen “ (Pras. v. j/ra lira- ,, fbrdern “). uttirali AV. 19, 32, 16,
,, aufbebend “ ? (Komm. iittarah !) : v. ud tr- ,, erboben
-tudd - : AV. 3, 25, la nt-tiidd- „ Aufstacbler “ (Augenblicks-
sondergott ; aus iit tudatu ebenda ; J. Zubaty, KZ. 40, 516), P. 3,
2, 35 arum-tiijda- ,, Wunden scblagend “ (ep, kl.) und vidhum-tvda-
,, den Mond verfinsternd “ (kb, Name Rabu’s), V. 1 zu P. 3, 2, 28
tilam-tuda- ,, Sesamkorner stampfend “ ; v. tiid- (Pras. tudd-)
,, schlagen “ {-tud- „ scblagend “ erst S. ep.).
V. turd- vielleicht in zwei Worter zu trennen : tu-rd- ,, stark “ zu v. tdvyas-
starker “ usw. (vgl. Ai. Gr.^ i, 22) und tur^d- ,, schnell “ von B. tvar- „ eilen “ (nicht
von V. tr~ „ hinubergelangen “) ; doeh vgl. v. tliya- „ stark Adv. ,, schnell “ und
H, Petersson, Zwei sprachL Aufsdtze (Lund, 1917), 62 ff. Wie verhalten sich dazu
V. dtura- ,, krank “ und v. ayidturd- ,, wohlbebalten “ ?
V. turvd- Name ; v. tiirv- (Pras. turva-) ,, iiberwaltigen “ ?
kl. tula- ,, die Wage im Tierkreis “ : ep. kl. tul- ,, aufbeben “.
-trpd - : RV. 4, 5, 146 a-trpd- ,, nicbt befriedigend “ : v. trp-
(Pras. trpd-) ,, befriedigend “ (vgl. v. -tfp- ,, sicb ergotzend “).
trsa- Divyav. fiir kl. trsd- „ Durst “ (Renou, Gramm, sanscr., 229) :
V. trs- ,, durstig sein “.
-tka- : AV. 2, 3, Ic ava-tk-d - ,, berabstiirzend “ : v. tak- ,, stiirzen,
fliessen “ (J. Wackernagel, KZ. 61, 190).
dasa- JB. 2, 183, nach Caland „ Bremse “, also ,, Beisser “ : v.
dams- (Pras. ddsa-).
U. ddpa- ,, Leuchte “ : AV. dip- ,, leucbten
-dughd - : VS. hhaga-dughd- ,, Anteilgewiihrer, Verteiler “ ; mit
anderm Akzent dughd- ,, Milchkub “ RV. 8, 50 (= Val. 2), 3d ; 10,
67, 6a, -dugha- (P. 3, 2, 70, Sonderregel) ,, (Milcb) spendend ; stromen
lassend “ v. hinter sukra- und sabar-, sonst nur Fern, -dughd- v.
hinter madhu- und su-, Samh. hinter kdma- und gharma-, ep. hinter
drorm-, Kas. zu P. 3, 2, 70 hinter artha - : v. duh- ,, Milch geben,
melken “ : Doch vgl. -duh- als Nomen agentis v. hinter go-, sabar-,
ghrta-, SV. hinter payo-, TB. S. hinter ghartna-, ep. kl. hinter kdma - ;
also ist -dugha- zum grossten Teil nur Erweiterimg von -duh- (vgl.
Ai. Gr., ii, 1, 222).
durd - ,, Erschliesser “ RV. I, 53, 2a, b : v, df- ,, bersten, zersprengen ** ?
-duha - : Mbh. dur-duha-, Spr. su-duha- „ schwer, leicht zu melken “ :
V. duh- „ melken “.
ep. -dusa- ,, verunreinigend “ : v. dus - ,, verderben “.
-dfsa- ,. aussehend “ (P. 3, 2, 60, und V. 1) seit TS. 7, 3, 17, 1
(Mantra’s) hinter 1-, kl-, td-, sa- neben dem altern v. -drs - ; vgb.
auch Lamnan, JAOS., 10, 489 f. ; fur Herleitimg von -dfsa- aus -dfs-
492
A. DEBRUNNEK
spricht auch das seit dem EY. belcgte und von P. vorgeschriebene
Fem. -dfsi-. Im Anschluss an diese Komposita ist auch v. su-ddrsa-,
U. dur-darsa- (vgl. Pali sii-dassa-) ,, leicht, schwer zu sehen “ zii ep.
{su-)dur-drki~ (Pali {su-)dud-dasa-) umgestaltet worden. Mit kom-
positionellem a v. TS. (a.a.O.) v'l-sa-drsa-, TS. (ebenda) su-sa-drsa-
(= V. su-sn-drk). AY. mndhu-sam-drsn- (v. sam-drs-) ; vgl. Ai. Gr.,
ii, 1, 108 ff.
ep. di/iita- (Y’hitney, ,,Eoots‘‘); v. dijiit- „leuchten‘‘.
-dm- : ptimm-dmm will Benfey, OOti. Ahh. 2.> (1879), iv, 2. S. 8. fnr puram-dardm
in dem uberzahligen Vers R\’. 8, 50 (61), 8r, schreiben (v. dr- ,, zersprengen “■) mit
Beriifuiig auf -dhra- nebcn -dharrt- (s. unten) ; andre suchen anders zu heilen (s.
Oldenberg z. St.).
kl. dvisa- ..Femd“ : v. dvis- (Priis. S. kl. dvisd-) ,,hassen“, vgl.
auch V. dvis-, f. ,, Feindschaft, Feind ”.
-dhni- : ep. kl. mahl-dhra- (Ganar. 8. 460 ; Yam. 5, 2, 36 ; Siddh.-
K.) neben ep. kl. wnffi-dhara- ,, Berg “ (,, Erdentriiger “) ; v. dhr-
,, tragen “. Ygl. auch Lex. lit-dhra- ,, Berg (mit dem kn- von Ai.
Gr., ii. 1. 83 ?). Y. ddhrd - ., diirftig. arm ; ep. kl. d dhr - bewahren " I
V. flhruvd' ,, fef^t == aw. drva-., ap. durura- gesund “ naeh H. Osthoff, Etym.
Parergur i, 119, \’on dru- ,, Baum mit Anschluss an die Wurzel dhr- ,, festhalten ”,
also Mischung aus denominativer und deverbativer Biidung.
-nnda- in ep. kl. apanuda - ,, vertreibend (P. 3, 2, 5 hinter sola-) :
V, nud- (Pras. nudct-) ,. wegsto.s.sen " ; doch vgl. auch U. ep. BhP.
-Had- .. wegstossend
y. piM- ,, Uamhirsc'h “ : v. pii- ., .schmucken " ?
-plda- : ep. Ixla-plda- ,, Olpresser, Olmuller ep. Irna-pnda-
,, das Zerdriicken wie Gras “ ; v. pld- ,, driicken “.
B. pidn- „ Falte, Ta.sche : kl. put- (Priis. putd-) ,, zerreibon u.a.“
B. p-dya- ,, Jauche, Eiter “ : B. pay- (Priis. pdya-) ,, stinken “.
kl. -pfira- ,, (sich) fiillend ep. das-pura- (Pali diip-pura-), BhP.
dur-d-pfirn-, Spr. su-puru- ,, .schwer, leicht zu erfiillen , kl. pfira-
,, Flut “ : AY. purdya- ,, fiillen “.
SB. prihdi- ,, flache Hand, Haiidlange TB. prt/ia-mdlrd- ,, Hand-
breite ep. prtJid- Eigenname : v. prath- ,, ausbreiten “. AY. 13,
2, 266 visvd'as-pitlaih ,, mit Hiindr-n auf alien Seiten “ fiir EY. 10,
81, 36, risvdtas-pdt .. mit Fiissen a.a.S.”
V. -prd- ,, erfiillend “ nicht zu Wurzel pf-, sondern zur Wurzelform v. prd- ; vgl.
auch y. -pra- ,, erfullend
v. priyd- ,. lieb ” = aw. fryct- (lies */riya-), germ. *frija- in ahd.
Frijd, aisl. Friag usw. ; v. prl- ,. erfreuou '' (P. 3, 1, 135 ; -iy- nach
Ai. Gr., i, 198) : vgl. auch v. -prl- ., sich erfreuend, liebend ".
ep. kl. hudha- (Kas. zu P. 3, 1, ].35) .. king ", als Eigenname B.
DEE TYPUS TUd1~ IM ALTINDISCHEN
493
ep. kl., AY. su-budha- „ wachsam a-budhd- SB. 14, 7, 2, 14 (= a-
budh- BAU. 4, 4, 11), kl., dur-budJia- ep., umr-budha- Lex. .spatkl.
(=v. wsar-6M(//?- „ friili wach “) : v. ,, wacheii
-brha- : Kaus. 75, 10, vi-brha- ,, das Losreissen : v. vi hrh-
(Pras. AY. brha-) ,, wegreissen “. Unklar Lalit. abrhu- oder arrha-
,, eine buddh. Gotterklasse
-bda- : EY. 7, 104, 17d upa-bdd- (= liaufigerein v. upn-bd-i-)
,. Getrampel “ neb.st durd-iipabdu- ,, weitliin schallend 7, 21, 2d ;
V. pad- ,, fallen (^\ alde-Pokorny, Vergl. Worterb., ii, 24, wo aw. und
gi'iech. Yerwandte verzeiclmet sind) oder von v. pad- ., Fuss “ (At.
Gr., i, 76) ?
-bruvd- : AY. 15, 13, 6c vrdtya-bmvd- ,, sick (falschlich) Yratya
nennend “, ep. brdhmana-brura- usw. : v. bru- (v. bruednti !)
,, sprechen ; vgl. Ai. Gr., ii, 1,183 (-uv- i, 198).
bhida- Pat. zu Y. 2 zu P. 3. 1, 135, ep. (su-)dur-bhida- ,, (sehr)
schwer zu sprengen " : v. bind- ,, spalten aber auch v. -bhid-
,, zerbrechend
ep. kl. bhuja - ,, Arm “, Y. 2 zu P. 3, 2, 5 mdla-ri-bhuja- Wurzeln
niederbiegend “ : v. bhuj- (Pras. hhijd-) .. biegen “.
bhrmd- „ Verirrung “ (?) RV. 7, 1, 22c ; 8. 50 (61), 12c (Oldenberg : ., sich rasch
tummelnd “ ? ) : ep. kl. bhram- ,, umhcTstreifeii
-blird- : v. an-ava-blird-nHlhas- ,, dessen GescLenke nicht zu
entreissen sind •' ; kl. ura-bhru- ,, Widder “ (nacli BE. „ AVolle
[*ura- — urnci-] tragend •'), daraus M. ep. kl. aurabhra- voin
Y’idder, Schaf stammend “ ; pra-ba-bhrd- Schleuderer “ Kath.
10, 9 (135, 12. 13). MS. 2, 2, 10 (23, 12. 13) immer im Zusammen-
liang niit vdjra- ,, Donnerkeil " (vgl. dazu EY. 1, 61, 12«, b ; 2, 30,
36 ; 5, 32, 7c, und bahhrir vdjrain ,, den D. tragend ■' 6, 23. 46) ; v.
66/- ,, tragen “, pra bhr- ,, sehleudern
-bJtva- : V. dbJiva- (AY. abhvd-) ,, ungeheuer, Unwesen “, v. vibhva-
,, ausgezeiclmet “ (neben v. vib/n'i-, v'lbJtu-, vibhean-, vibhvdn-) : Erwei-
terung aus v. -bhil- ,, seiend “ I
-rnisd- : v. a-ni-mim- (und d-ni-mixat-) ,, wachsam " : v. m
misati ,, schliesst die Augen “. Ygl. auch die adverbiellen v. d-ni-
mis-am und -a, die ein J7om. act. voraussetzen ; also ist -misa-
Erweiterung des athematischen Nominalstamms im Komposituni
(s. Ai. Gr., iii. 323).
S. -nfila- (Whitney, „ Eoots “) : v. mil- (Pras. B. mtla-) „ die
Augen schliessen “.
-muca- : Y. 2 zu P. 3, 2, 5, Ganar, 8, 460 nakJia-muca- ,, die Finger
494
A. DEBEUNNEE
loslassend “ : v. muc- (Pras. miicd-) ,, loslassen Aber AV. dmuci
Name einer Damonin gehbrt zu v. -7nuc- ,, loslassend, befreiend
-mudd- : AV. hasd-tnudd- ,, frohlich lachend nacb Kas. zu
P. 3, 2, 5 (vgl. Ai. Gr., ii, 1, 83) auch in AV. ku-muda- ,, weisse
WasserliUe “ : v. rniid- „ froblicli sein aber auch v. 7nud- ,, Lust “.
mur-mura- ist zweifelhaft ; s. Ai. Gr., i, 23.
mrksd- „ Striegel “ RV. 8, 55 (66), 3a ; v. 7nrks- (Pras. rnrksd-)
,, striegeln “.
-mrja- : tunda-pari-mrja- ,, sich den Bauch streichend “ P. 3, 2,
5 (V. 1 : nur im Sinn von ,, trag “), Lex. ; v. jxiri tnrj- (Pras. tnrja-
AV. S. kl.) ,, rings abreiben “.
Lex. 7nrja- ,, eine Art Trommel “ : Dhatup. tnrj - ,, tonen “.
Kath. mrda- ,, gnadig “ ; v. tnrd- (Pras. mrdd-) ,, gnadig sein
S. -mrda- (Whitney „ Roots “) : v. 7nrd- ,, zerdriicken “.
-mrdhd- : vi-tnrdhd- ,, den Verachter abwehrend “ RV. 10, 152,
2b, AV. TS. ist Erweiterung von vi-tnfdh- VS. usw. (At. Gr., ii, 1, 109),
vgl. V. mfdh- „ Verachter “.
-mrsd- : VS. 16, 36 pra-mrki- „ antastend “, BhP. n-mrsa-
,, Priifung “ (= ep. kl. vi-marsa-) : v. mrs- (Pras. mrsd -) ,, beruhren “.
-»7wa- : pari-rfUd- „ Anteil “ (?) BV. 1, 187, 86 : v. nos- „ erreichen “ (vgl. v.
dipsa- „ Anteil “) ? Vgl. Oldenberg z. St.
V. yugd- ,, Joch “ = np. juy, lat. jugum, gr. ^vyov, got. juk, abg.
{go, heth. yukan: v. yuj- ,, anschirren “.
-yuja- : a-yujd- „ ohne Genossen “ RV. 8, 51 (62), 2a, A^vGS.
(nebst ayujdksara- ParGS.) mit kompositionellem a zu v. yuj- -yitj-
,, verbunden, Genosse “.
-yudha- : v. d-yudha- n. ,, Waffe “ ebenso zu v. yudh- ,, Kampf “,
-yudh- ,, kampfend “.
V. yiipa- „ Opferpfosten “ ; v. yup- „ glatten “ ?
-risd- : AV. 7mghd-risd- Behvort von Pflanzen : v. ris- ,, Schaden
leiden “ ? vgl. auch v. ris- ,, Schaden, Schadiger “.
rued- ,, licht “ VS. 31, 20 (ruedya wohl nur spielerische Variation
von ruc-dm 31, 21), MS. 4, 9, 5 (125, 11) (ruco ’si ebenso neben allerlei
Formen der Wurzel rue-) : v. rv.c- „ leuchten “ (auch Priis. Ptz.
ruc-dnd-). ep. Lex. rued- ,, Gefallen ” ist Erweiterung von v. rue-,
f. ,, Glanz, Wohlstand “.
ruja- ,, brechend “ AV. 16, 3, 2 (?), v. hinter valani- und d-, Ragh.
P. 3, 2, 31 hinter kulam-ud- : v. ruj- (Pras. ritjd-) ,, brechen “.
kl. -ruda- (Whitney, „ Roots “) : v. rud- (Pras. S. rudd-)
,, jammern “, vgl. auch AV. -rud- „ jammernd “.
-rudha- : v. d-go-riidha- „ die Kiihe nicht fiir sich behaltend “
DEE TYPUS TVDA- IM ALTINDISCHEN
495
(s. W. Neisser, Znm Worterbuch des Rgv. ii, 1930, 9) : v. rudh-
,, zuriickhalten
-rudha- ,, wachsend “ (v. rudk- ,, wachsen “) nur in virudhayidyn AV. 6, 21, 2b :
Reimbildung zu bhesajdndm in a ; sonst v. vtrudh- „ Gewachs, Kraut
-ruha - : ep. kl. -d-ruha- ,, besteigend dur-d-ruJia- ,, schwer zu
besteigen kl. (Pat. zu V. 2 zu P. 3, 2, 5) samsi-ruha- ,, im AVasser
wachsend ; v. ruh- (Pras. ep. kl. ruha-) ,, wachsen “ ; aber vgl.
auch V. ruh - ,, Wuchs, Trieb seit RV. -ruh- ,, wachsend “.
lih'ha- ,, ritzend “ Siddh.-K. 2897, vi-likha- Kas. zu P. 3, 1, 135 :
AV. likh- (Pras. Itkhd-) ritzen “.
-Uga - ; AV. 5, 13, 7a dligl- und viligi- als Schlangennamen, wohl
Fem. zu -Uga- : ep. kl. (d-)lihg- ,, umschlingen “.
-Upa- : KaA zu P. 3, 1, 138 pra-lipa- : v. lip- ,, bestreichen “.
-lisa- : V. ku-lisa- „ Beil “ (,, schlecht abrupfend “ ? s. Ai. Gr.,
ii, 1, 83) ; v. ris- (Pras. risa-) ,, zerreissen “.
-liha - ,, leckend “ ; P. 3, 2, 32 vaham-liha- ,, die Schulter leckend “
und abhram-liha- (ep. kl.) ,, die Wolken erreichend “, Lex. go-liha-
,, eine Pflanze “ : v. lih- (B. Pras. liha-) ,, lecken vgl. auch kl.
lih- ,, leckend
-luncd- : VS. ku-luiicd- ,, Ausraufer “, kl. a-luncd- ,, nicht
nipfend “ (?) : ep. kl. lunc- (Pras. luhca-) ,, raufen “.
kl. -vida- ,, wissend “, ep. (su-)dur-vida - ,, (sehr) schwer zu wissen “ :
V. rid- (Pras. AV. vidd-) ,, wissen “ ; vgl. v. -lid - ,, wissend “. Unver-
standlich AV. 19, 22, 18 vidagand-, unklar RV. 5, 45, la vidd, Pp.
viddh ; der Eigenname vida- (S. kl.) ist bida- zu schreiben.
Lex. vidha- ,, Bohrer “, V. 4 zu P. 3, 3, 58, und Pat. dazu d-vidha-
,, Bohrer “ : v. vyadh- (Pras. v'ldhya-) ,, durchbohren aber auch v.
{hrdayd-) v'ldh- ,, (das Herz) verwundend “.
-visa - : R. dur-visa- schwer zu betreten “ : v. vis- (Pras. visd-)
,, betreten “.
visa- : RV. 8, 19, 11c ; 10, 109, 5a vtsah nicht von *risa - ,, Diener
sondern von vis-, f. ,. AVerk “ (s. Oldenberg zu 8, 19, 11c). Mbh.
dur-visa- als Beiname diva's, eigentlich ,, mit deni man schwer fertig
wird “ (?) : v. vis- ,, geschaftig sein “.
-Visa- ; V. pdd-bida-^ VS. pdd-i Tsa- ,, Fussfessel “ : zu lat. vincire ? (s. Ai. Gr.^ i,
183, wo padvimsa- nach Fr. Edgerton, JAOS.y 51, 170, zu streichen ist ; Walde-
Pokorny, Vergl. Worterb., i, 234).
V. vrdhd- (auch als Hinterglied) ,, erfreuend. Forderer “ : v.
vrdh- ,, vermehren, erfreuen vgl. auch v. I'jrdh- ,, Forderung “,
-rfdh- ,, sich freuend “. AV. 2, 13, 5c, su-vfdhd Instr. von su-vfdh-.
ep. kl. rfsa- (Akzent nach P. 6, 1, 203) „ Stier “ u. a. nicht aus der Wurzel v.
tT5- ,, regnen, benetzen sondern aus der Hintergliedform -vrs-d- (AV.) von v.
vfsari- ,, Stier
496
A. DEBRUXNER
V. vrd- m. ,, Schar “ (BR., angeblich von v. vr- „ umschliessen ") ist durch vra-,
f. „ lockendes Weib “ zu ersetzen (Grassmann, Worterbuch, s.v. ; Pischel, 1 ed. Stud.,
313 flf.).
ep. kl. vr'ida- (haufiger vrida-) „ Scham : ep. kl. vrld- (Pras.
vrtda-) ,, sick schamen
-sasd. : SB. 10, 5, 2, 5, uktha-sasdm (Akk. Sg. m.) enthalt nicht Tiefstufe von
Sams-, sondern ist irgendwie eine Umgestaltung des v. uktha-Sas- (junger -sds-)
„ den Spruch hersagend das aus der urspriinglich nasallosen Wurzel sas- (s. Ai. Or.,
iii, 250) gebildet ist.
V. sit'd- = V. sera- ,, lieb, vertraut “ (idg. *kei-uo-, germ. *heiwa-
und *hhva- ; Walde-Pokorny a.a.O., i, 359).
kl. -sisa- (UTiitney „ Roots : v. sts- ,, iibrig lassen
-Slla- als Nom. ag. mit Vordergliedbetonung lebrt V. 7 zu P. 3. 2, 1, dazii Pat.
rndtrisa-Slla- „ an Fleischnahrung gewohnt *■. Xicht von Dhatup. silati noch vom
Denominativ S. kl. Sllaya- (so das Vartt.), sondern Bahuvrihi mit VS. slla- „ Gewohn-
heit “ {.Ai. Or., a, 1, 218 f.).
hied- RV. 10, 26, 6b, „ leuchtend “ (? nack Oldenberg z. St.
unklar) : v. sue- „ leuchten “ ; vgl. jungaw. suea-, sfica- „ sekend,
licht
ep. kl. subha- „ hiibsck “ : v. subh- „ schmiicken vgl. auch v.
hibh- ., Schonheit “.
-Suva- ( ; V. sii- ,, schwellen “) angeblich in ahl-mvah RV. 10, 144, 3c (Damonen-
bezeichnung) : dies ist aber vielmehr Akk. PI. von v, ahl-su,- ,, von Schlangen
schwellend ; s. Ai, Gr„ ii, 1, 222, und Oldenberg z. St.
V. susd~ ,, Kraft, rautig “ nicht aus der AVurzel v. svas^ sus- ,, blasen , sondern
^-Enveiterung der W’^urzel v. ,, schwellen
-srmbhu - : v. ni-srnMd- „ sicker auftretend “ : ep. kl. srambh-
„ vertrauen “.
-Sratha- : P. 6, 4, 29, lebrt fur die unbelegten pra-Sratha- und hima-Sratha-
.Schwund eines Binnennasais ; vgl. ep. kl. Slatha- ,, locker . Poch kennt die vor-
klassische ,Sprache ausser dem inli.xalen dnu sriUhati T8. 6, 1, 9, 7 nur nasalloses srath-.
-sajd- : V. eakraiu-d-sajd- ,, das Rad hemmend ; v. sanj-
(Pras. d sajdrni) ., ankaften, anheften “ ; s. Ai. Gr., ii, 1, 183.
Unklar siva- Pat. zu V. 2 zu P. 3, 1, 135.
V. siida-, m. ,, Siissigkeit “ : v. srdd- sud- ,, siiss, angenekm sein “.
Aber TS. 1, 3, 3, 1 (Mantra) havyn-sudn- ., die Opferspeise bereitend “
(die Paralleltc.xte 3IS. Kath. PB. lesen -siidana-) ist Themati.sierung
von V. havya-sud- .
ep. stubha- Name eines Agni : v. stubh- „ preisen vgl. auck v.
stubh-, f. ,, jauchzender Ruf.“
-spira- ( : v. spr- ,, losmachen “) in RV. 5, 43, 14* raspird- ? Doch ist dieses wie
das amscheinend gleichbedeutende rdspind-, 1, 122, Ad, unerklart ; vgl. At. Or., i,
23, iii, 215, Oldenberg zu 1, 122, 4.
-sprsa- : Hariv. du(h)-sprsa- ,, unangenehm anzufassen “ ; v.
sp/®- (Prils. sprsn-) ,, anfa.ssen vgl. aber auch v. -sprs- „ berukrend “.
DEE TYPERS TUDA- IM ALTINDISCHEN
497
sphiga- Benfey, Vollst. Gr,, 135, § 368, 1 A 4 aus Gana karna, wo Bohtlingk in
seiner Ausgabe des Panini sphij- liest. apa-sphigd- lehrt P. 6, 2, 187 : Wurzel ?
ep. kl. spJiuta- ,, aufgebliiht, offenbar “ : B. sphiit- (Priis. sphutd-)
,, bersten
-sphura - : AV. 1, 2, 36 anu-sphurd- ,, schwirrend “ (vom Pfeil),
RV. 6, 48, 11c dn-apa- sphura- ,, nicbt wegsclinellend : v. sphr-
(Pras. sphurd-) ,, wegstossen aber auch RV. 8, 58 (69), 106. c apa-
sphur- dn-apa-sphur- ,, (nicbt) wegschnellend
kl. sphurja- eine Pflanze, ein Raksasa : AV. sphftrj- (Pras. sphurja-)
,, brummen “.
SB. sydda- ., das Fabren “. P. 6, 4, 28 ,, Gescbwindigkeit “ (mit
Scbwund des Nasals), dazu Kas. go-syada- asva-syada- : v. syand-
(mit Aor. d-si-syadat) ,, forteilen Vgl. aucb AV. sanisyadd-
,, fliessend “ zu v. sdnisyadat, ferner v. sasydd-, f. ,. eilender Strom “,
havana-sydd- ,, auf den Ruf berbeieilend “.
V. sruvd- ,, Opferkelle “ wohl zu v. sru- ,, fliessen “, vgl. v. sruc-
,, Opferlbffel “.
AV. svajd - ,, Viper “ : svauj - ,, umscblingen “ (AV. pdri-svanjalya-,
Pras. V. svdja-).
B. hira- „ Band “, Satnh. hird- ,, Ader “ : v. hr- ,, nehmen “ ? (s. .4!. Gr., i, 23).
hidd- ; RV. 8, 12, 19a hidd(h) enthalt kein *hidd- ,, Zupfer, Mahner (BR.) zu
V. hid- ,, zumen, zupfen “, sondern ist Abl. von kid- „ Zorn (Oidenbcrg z. Bt.).
hurd- : RV. 4, 3, 13a kurdh nach Grassmann {Worterb., s.v.) ., Riicher “ von
Wurzel V. hvr- ,, beugen, storen “ ; doch eher nach BR. und Oldenberg z. St. Adverb
,, auf krummem Weg “ (vgl. .4i. Gr., i, 23).
-hra- „ rufend Pat. zu V. 1 u. 2 zu P. 3, 2, 3, gehdrt zur Wurzelform B. hid-,
nicht zu V. hu-.
*-zda- : V. nidd- ,, Rubeplatz, Lager “ aus idg. *ni-zd-o- (Ai. Gr.,
i, 76), vgl. lat. nidus, abd. nest usw. : v. sad- ,, sicb setzen “.
[Nacbtrag ; Uber die alten Herleitungen wie v. k.slrd-. n. .. Milcb
aus V. k.sar- .. fliessen u. dgl. (so nocb Charpentier. Monde oriental
18. 1924. 35. und Renou. Gramm, sanscr. 214) s. Ai Gr.. i. 25. 27
und oben jlra-.']
Wer das vorsteliende Material durchgeht. sieht sofort. dass ein
einbeitlicber Bildungstypus nicbt vorliegt und dass sick diese Worter
weder an Hiiufigkeit nocb an Regelina.ssigkeit mit den Verbalnomina
auf a mit Guna der Wurzel (z. B. -lard-, -cartd-, veda-) messen konnen.
Doch beben sicb drei Gruppen heraus :
1) Am deutlicbsten i.st die Beziehung zu Prasenshildungen mit
Tiefstufenvokal und Suffixbcionung, d.h. zu Prasentien der 6. Klasse.
So V. -i.sd-, -ubjd-, -trpd-. -mi.sd-, -mrhm-, -rujd-, -lisa-, -sphurd - ;
Samh. -lihidd-, -gird-, -tudd-, -mrdd-, -mrsd - ; B.-S. Tsd-, -uksa-.
498
A. DEBRUNSER —
-ujjha-, uncM-, -jusa-, -tird-, putd-, -hrha-, -visa- ; ep. guha-, -nuda-
-bhuja-, -ruJia-, -vida-, -sphiita-, -sprsa- ; kl. Jcsipa-, dcisa-, -muca-,
-mrja-, -nida-, likha-, -liha-. Solche Beziehiingen von Nomina agentis
auf a zu thematisclien Prasensstammen sind langst bekannt : die
Nomina agentis steken in ilirer Bedeutung den Partizipien sehr
nahe, ausserdem fielen im Ai. wegen des Zusammenfalls von idg.
e und o die Nomina agentis vom Typus -vartd-, coda-, usw., mit den
zugehorigen Prasensstammen der 1. Klasse volKg zusammen. Vgl.
Ai. Gr., ii, 1, 178 ff. Seltener sind Bezieliungen zur 2. Klasse : -dugJia-,
-duha-, -bruva- {-bhra- ?), -mrja-, -liha-. Sie erinnern an die Nomina
agentis auf a aus (athematischen) Intensivprasentia : v. vevijd-
„ auffahrend “ zu v. ve-vij-ana-, v. ku-namnamd- ,, sich ungern
beugend “ zu v. mim-nam-ili, usw.
2) Auffallend sind auf den ersten Blick die Bezieliungen zu
Praseutia der ersten Klasse mit prosodiscb langer Wurzelsilbe, d.b.
mit i, u, r vor mebrfacher Konsonanz oder mit ?, u (f) vor einfacbem
Konsonant (oder vor Konsonantengruppe). Das Staunen weicbt
aber, wenn man beachtet, dass diese Wurzeln entweder liberbaupt
keiuen oder fast keinen Guna kennen, auch in sonstigen Guna
erfordernden Verbalformen und Ableitungen. Die von solcben
Wurzeln abgeleiteten steigerungslosen Nomina auf a sind also den
gunierten gleichwertig. Daher sind aucb in dieser Gruppe die
Nomina actionis und die Simplicia verhiiltnismassig viel haufiger
als in Gruppe 1) und 3). da ja die gunierten Bildungen ebensowohl
Nomina actionis wie Nomina agentis bilden (z. B. v. savd- ,, Antrieb
aber ap-savd- ,, Wasser spendend “) und die Nomina actionis viel
haufiger Simplicia sind als die Nomina agentis. Nach dem Vorbild
der Gruppe 1) sind die moisten endbctont trotz der Barytonese des
Prasensstamms. Die Boi.spiele die.ser Gruppe sind ; v. -mkhd-, -krldd-
jivd-, tun'd- ; Sarnh. -ukd- (1), -luhvd- ; B.-S. isd-, iiiichd-, kurda-,
dipa-, puya-, -mlla- ; ep. inga-, Ira-, -uha- (2), kuja-, -plda-, vrlda- ;
kl. -Iksa-, iha-, guhja-, giiha-, jrmbha-, tula- (Kausativ ep. kl. tulaya-
neben tolaya- !), sphurja-. Auch in der 6. Priisensklasse komineu
Wurzeln von diesem Bau vor, daher v. -ubjd-, Samh. mrda- (das r
dieser Wurzel wird im EAh lang gemessen !) ; B.-S. -uk.sa-, -ujjha-,
uhcha- ; ep. guha-, ebenso bisweilen vor einem prasensbildenden -aya-,
daher v. kutsa-, AV. gulphd-, ep. -pura-. Endlich darf man auch v.
-sajd- und Samh. svajd- hierher rechnen, well das a im zugehorigen
PriLsensstamm betont ist, obwohl es auf idg. n zuriickgeht.
3) Klar sind auch die Beziehungen zu den athematischen
\\ urzelnomina ; in manchen Fallen ist der a-Staram geradezu als
DEK TYPUS TUdI- IM ALTINDISCHEN
499
Erweiterung des atheniatischen zu betrachteii. Hierher gehoren ;
V. -isa-, -gJind-, -trpd-, -ditgha- {-dugJid-), -bhva-, -misd-, -mrdhd-,
-yujd-, -yudha-, vrdhd-, -sphiird- ; Samh. -jum-, -tudd-, -dfsa-, -budJia-,
-muca-, -mudd-, rued-, -siida- ; B. Jsd-, -ulesa-, sydda- ; ep. -mida-,
-bhida-, -ruha-, -vida-, -sprsa- ; kl. dvisa-, -muca-, -rnda-, -liha-, -vidha-.
Die Erweiterung mit a tritt bekanntlich besonders gern am Hinteiglied
eiaes Kompositums auf (vgl. Ai. Gr., ii, 1, 222), aber auch am einfachen
Wort (vgl. ebenda iii, 319 ff.). Das Schwanken des Akzents erklart
sich daraus, dass diese Bildungen entweder die Ton.stelle des
atkematischen Stamms beibebalten oder — was weit haufiger ist —
sich an die Falle von 1) anschliessen : eine ganze Reihe dieser Bildungen
musste ja auch unter 1) erwahnt werden, well eben manche Wurzeln
ebensowohl ein tiefstufiges athematisches Wurzelnomen wie ein
Prasens der 6. Klasse bilden.
Bemerkenswert sind als Storenfriede die potentialen Zusammen-
setzungen mit su- und duh-. Diese Bildungen haben normalerweise
Guna, gehen aber gelegentlich trotz der Bedeutungsverschiedenheit
in den Tj'pus h(da- iiber. Diese neuen Bildungen tauchen erst im Epos
auf (sind aber da besonders beliebt), und mehrere einzelne Bildungen
sind offenkundig j Unger als die Gunabildungen aus derselben Wurzel ;
ep. {su-)dur-vida-, aber su-veda- ,, leicht zu erlangen “ schon v., dur-
veda- ,, schwer zu flnden “ schon SB. ; -drsa- und -darsa- s. in der
alphabetischen Liste ; duh-sprsa- nur Hariv., aber sonst ep. kl.
duh-sparsa-. Die ubrigen Beispiele s. unter -uha-, -duha-, -pura-,
-budha-, -bhida-, -ruha-, -vida-, -visa-.
Aus dem Bestand der drei Listen ist es verstandlich, dass Panini
3, 1, 135 als allgemeine Regel die Bildung mit Suffix (k)a, d.h. a
ohne Steigerung der Wurzel, aus Wurzeln mit kurzem oder langem
antekonsonantischem i, u, r lehrt. Aus dem Prasensstamm leitet er
nur die Bildungen mit Akkusati\dbrm des Vordergheds ab (Suffix
Qch)a[s) ) : 3, 2, 31 Icularn-ud-ruja-, 32 vahani-Uha-, abhrani-liha-, 35
vidhurn-tuda, arum-tuda- ; hier war eben der Hinweis auf partizipahn-
liche Funktion des Hinterglieds durch die Kasusform des Vordergheds
gegeben.
Bemerkenswert ist, dass P. das Suffix {k)a auch fur die Bildungen
aus Wurzeln auf d ansetzt (3, 1, 135. 136 ; 3, 2, 3. 4. 6. 7. 77). In
der Tat enthalt v. -dd- ,, gebend “ die schwachste Stufe -d- der Wurzel
dd- vor betontem Suffix d wie -tudd- die von tud-, tod- ; auch verhalt
sich -d-a- als Erweiterung zu -dd-, -d- ,, gebend “ ebenso wie -yuj-d-
zu -yuj-.
Beziehung eines Suffixes auf einen Prasensstamm ist ebenso
500
DER TYPUS TVDA- IM ALTINDISCHEN
sekundar wie die Erweiterung eines athematischen Stammes ; es ist
daher kein Wunder, dass von den Beispielen unter 1) und 3) ausser
-ghna- und jivd- keines eine ausserindische Entsprechung hat. Und
zwar gilt bei -gJina- die (aw.) Parallele nur dem Abstraktum mit
nominalem Vorderglied, also einem Tj’pus, in dem die Erweiterung
mit dem Themavokal aus der Grundsprache stammt : es handelt
sicb streng genommen nicht um Komposita mit -ghna~, sondern um
Bildxmg eines substantivischen (abstrakten) Neutrums aus dem
zusammengesetzten Nomen agentis auf -han- (vgl. H. OsthoiF,
Sprachwiss. Abb. herausg. von L. v. Patrubany, ii. 119 f. ; Brugmann
IF. 17, 358 f. und Grundriss,^ ii, 1, 156 ff.). Und jivd jtvati setzt
ein idg. Wortpaar fort, dessen lautliche Gestaltung und gegenseitiges
Verhaltnis im Idg. ganz einzigartig dastehen. Zu einem idg. Typus
gehoren auch die Bildungen mit volligem Schwund des Yokals (idg.
e/o) zmschen Konsonanten : (-Jcra-), (-gnm-), -grd-, -tkd-, (-dra-),
-dhra-, (-bda-), -bhra-, -msa-, *-zda-, vgl. oben unter -zda- und ferner
griech. veo-yvos ,, neugeboren hi-^po? ,, Wagenkorb “ (,, mit
zwei Tragern, d.h. Seitenlehnen “), \a.t. pro-brum ,, Vorwurf Die
Worter, die sonst ausserindische Verwandte haben, sind ohne
Beziehungen zu Prasensstammen : *grdhd-, priyd-, yugdi-, sivd-,
sind auch nicht Nomina agentis und nicht (wie fast alle in der Liste
1) aufgefiihrten Worter) auf Hintergliedstellung beschriinkt, sondern
einfache Al'orter. Ein athematisches Wurzelnomen hat nur priyd-
neben sich ; aber v. -prl- ist akti\’isch, priyd- passi\'isch, es sind also
zwei von einander unabhiingige Bildungen. Die hochst sonderbare
Bildung idg. *iug6m erklart H. Jacobsohn (DLZ., 1912, 2786) ganz
plausibel aus einem kollektiven Plural *iugd ,,Joch und Zubehor
der zu einem IMaskulinum *iengo- (= v. ydgn-) ,, Anschirrung, Ge-
schirrstuck gehorte {yugd- als Mask, nur bei Le.xikographen, griech.
t^vyos heisst ,, IVagebalken, Joch “). Ausserindische Bildungen
von der Art wie -tuda- sind selten ; abg. po-c'du ., Aufziihlung “ u. dgl.
s. F. de kSaus.sure, d/c'/a. 228 Anm. = Hvrueil des pnbl. scienl., 214 f.
Anm. 1, Brugmann, Grioidr.,- ii, 1, 155, Stolz-Leumami, Lat.-Gr.^,
202. Lat. -dims {cnusi- usw.) und -ficus {muni- usw.) sind jiinger
als -dex (ju- usw.) und -fex {arti- usw.), s. Ernout-Medlet, Diet,
etym. 256, 307 f.
Der Bedeutung nach sind diejenigen, die Beziehungen zum
Priisensstamm oder zu den athematischen Wurzelnomina haben, in
der Regel Nomina agentis. was nach dem oben Gesagten verstandlich
1 st ; warum die zu Priisentia der erslen Klasse gehorigen eine Ausnahme
machen, ist oben S. 498 gesagt.
The Prakrit underlying Buddhistie Hybrid Sanskrit
By Franklin Edgerton
A SANSKRITIST who reads for the first time a Buddhistic Sanskrit
text such as the Saddharmapundarika is struck at once by
peculiarities of vocabulary and style which differentiate it from normal
Sanskrit. If he limits himself to the prose parts, ignoring the verses,
he will rarely encounter forms or expressions which are definitely
ungrammatical, or at least more ungrammatical than, say, the Sanskrit
of the epics, which also violates the strict rules of Panini. Yet every
paragraph will contain words and turns of expression which, while
formally unobjectionable (if, perhaps, non-Paninean), would never
be used by any non-Buddhist writer. If our Sanskritist is also familiar
with Pali, he will soon notice that many of these words and turns of
expression are identical, nnUatis mutandis, with Pali words and turns
of expression. For example, in SP., 76, 10 (I refer to page and line
of the Kern-Yaniio edition of the Saddharmapundarika), and often,
dtmabhdva occurs in the sense of ” body ". The word is a quite normal
and innocent-appearing Sanskrit formation, and occurs, e,g., in the
^vetasvatara Upanisad 1, 2, meaning “ existence (or reality) of the
self (soul) ; in the meaning ‘‘ body ", however, it occurs only in
Buddhistic Sanskrit, but there quite commonly. Now it cannot be
accidental that its phonetic equivalent in Pali, atlabhdva, has precisely
this meaning. Again, all readers of Pali are very familiar with the
common expression yena . . . lena . . . , “ where (someone or
something was), there (someone else went)." In themselves, yena
and tena are perfectly normal Sanskrit (as well as Pali) forms ; but
this use of them, I believe, is not known except in Pali and Buddhistic
Sanskrit, though frequent there. These are characteristic examples
which could be multiplied many times, as all students of the field are
well aware.
Such students also know, of course, that this is by no means the
whole story. In the cerses of such works as the Saddharmapundarika
or the Lalitavistara (and in the prose of, e.g., the Mahavastu. which
in this respect is unusual ; much more rarely in the prose of most
other works, at least as presented in our editions), there also occur
many forms which are unknown to Sanskrit grammar, of anv period.
502
F. EDGERTON —
aad would be felt as barbarous and impossible in any genuine or
“■ normal ” (non-Buddbist) Sanskrit work. They are, in brief, middle-
Indic ; in a broad sense, Prakritic. This fact, together with those
mentioned above, led some scholars of a generation or more ago (such
as Childers) to the not unnatural supposition that these Buddhistic
Sanskrit works were translations, or re-workings, of Pali originals.
More careful study of Pali itself, and of other relevant materials,
has shown that this hj’pothesis does not fit the facts. The striking
linguistic resemblances between Pali and Buddhistic Sanskrit do not
indicate any direct relation between the two dialects, or between the
literary works composed in them. But the relation, though indirect,
is nevertheless certain. Both contained originally texts which were
based on canonical texts composed in an earlier dialect, Prakritic
in character, in which there must have existed at one time a consider-
able body of (perhaps only oral) Buddhist literature. Neither the
Pali nor the Sanskrit Buddhist canon is “ original ”, nor is either based
on the other ; both contain, or once contained, essentially (in their
older parts) translations or recasts of compositions in that older
Prakrit. As time went on, both languages were then used in original
compositions (most of our actually extant Buddhistic Sanskrit texts
are, in fact, original, rather than translations or re-workings) ; but in
such a way that the traditional link with what we may call the proto-
canonical Prakrit was not wholly broken. At least in vocabulary,
and (particularly on the Sanskrit side) for a long time also in morphology
and even phonology, Buddhist writers, both northern and southern,
used idioms which were clearly under the influence of a linguistic
tradition stemming from that protocanonical Prakrit.
The fact that Pali is itself a middle-Indic dialect, and so resembles
the protocanonical Prakrit in phonology and morphology much more
closely than Sanskrit, makes it harder to trace such influences in it.
Yet, as Professor Sylvain Levi has shown,* Pali is not free from them ;
^ See his brilliant and important article of 1912, JA., Ser. 10, vol. 20, pp. 495-512.
I hope that Professor Levi would accept my formulation of the matter as above,
which I think differs little in principle from his, though he uses the term “ pre-
canonical ” rather than “ protocanonical ”, meaning, I take it, antecedent to the
historically known Buddhist canons. Since 1 think (and I presume the great French
savant would agree) that a“ canon ” in some sense doubtless existed in that language,
I prefer “ protocanonical ”, with Professor de la Vallee-Poussin {Indo-europeens et
Indo-iranienSf p. 202). The most important bibliographical references on the subject
will be found in these two places and in J. Mansion, Esquisse d'une histoire de la langue
sanscrite (1931), pp. 105-9, where will also be found interesting speculations as to
the manner of development of the curious “ Buddhistic Sanskrit ” dialect.
THE PRAKRIT UNDERLYING BUDDHISTIC HYBRID SANSKRIT 503
for instance, it now and then presents forms with loss of intervocalic
mutes, or sonantizing of intervocalic surds, contrary to the laws of
the Pali language. Levi has also shown that similar traces of this
protocanonical Prakrit can be detected in occasional words and
phrases occurring in the Asokan and other early inscriptions.
It is, however, in Buddhistic Sanskrit that we find the clearest and
most extensive evidence. Quite naturally ! For when Buddhist
monks began to adapt the language they used to the “ respectable ”
language of the Brahmans, any imperfections in the adaptation would
necessarily show up much more glaringly, than when they simply
turned it into another Prakrit (such as Pali, in essence, was) ; because
the linguistic gap between the two media was far wider. A relatively
early stage in this adaptation is represented by the verses of, say, the
Saddharmapundarlka. Here every line shows evidence of Prakrit
influence ; and that too not only in vocabulary, but also in phonology,
and especially in morphology'. As time went on, the tendency was
to approximate more and more the forms of standard Sanskrit, until
finally almost the only remaining trace of Prakrit consists in the peculiar
Buddhist vocabulary. (It should be emphasized, however, that this
vocabulary is itself evidence of appurtenance to a separate linguistic
tradition, quite distinct from “ standard ” Sanskrit. For it is not
merely a question of technical terms relating to religion, but very
largely of terms of every-day life. They can be explained only as
marks of a distinct language.) There seem to be reasons for assuming,
in general, that the more Prakritic a text looks, the earlier it is. To
be sure this cannot be taken as a hard and fast rule. What is certain
is, that nearly all Buddhistic works in Sanskrit (at any rate, until a
late period) belong to a continuous and broadly unitary linguistic
tradition ; their language is a thing separate from the tradition of
Brahmanical Sanskrit, and goes back ultimately to a (semi-) Sans-
kritized form of the protocanonical Prakrit. The number of Buddhist
writers who stood outside this tradition, that is who wrote in what is
virtually standard Brahmanical Sanskrit, seems to have been very
small. We may guess that it was limited to converts who had received
orthodox Brahmanical training in their youth, before adherence to
Buddhism. Asvaghosa is an example of this exceptional type. His
Sanskrit can probably not be distinguished from that of Brahmanical
writers in phonology or morphology, and only to a slight extent, if
at all, does he make use of the peculiar Buddhist vocabulary. Now
it is “ taken as certain that he was of Brahman family, and had enjoyed
504
F. EDGERTON
a thorough Brahmanical education before he went over to Buddhism ”
(Winternitz, Hist. Ind. Lit., vol. ii, 1933, p. 257).
It seems to me no exaggeration to speak of this hybrid Sanskrit of
the Buddhists as a language, in its own right. Not a vernacular, of
course ; a literary language ; an artificial language, if you like. I
grant, also, that it appears in various markedly different phases,
distinguished chiefly by great differences in degree of Sanskritization
(approach to normal Sanskrit in phonology and morphology). But
these phases are aspects of a unitary tradition, connected with each
other by direct lines. To trace these lines in detail would be to construct
a relative chronology of the Buddhist Sanskrit literature. It must
be admitted that this is at present impossible. Perhaps it will never
be possible. Nevertheless the underlying unity of linguistic tradition
seems undeniable.
It is signalized, first, by the peculiar and persistent vocabulary
referred to above. Boehtlingk included some of it in his great Sanskrit
dictionary (how many words, or special meanings of words, are there
recorded only from Buddhist works !) ; but perhaps the larger part is
not included in any Sanskrit dictionary. And. in strict linguistic
logic, it should not be there ; that is, unless we stretch the meaning of
“ Sanskrit ". The fact that Pali contains so large a proportion of these
words seems to prove that most of them must belong to the special
vocabulary of the protocanonical Buddhist Prakrit. (It may be noted
in passing that they are, in general, not common Prakrit ” ; relatively
few, I believe, will be found in Prakrit guise in Sheth's Prakrit Dic-
tionarg, for instance.) They characterize all periods of Buddhist
(hybrid) Sanskrit. We need a special dictionary of this language.
It is signalized, secondly, by peculiarities of syntax and style.
I recall the gena . . . term construction (above) ; or the use of third
person singular verbs with subjects of any person or number, which
goes beyond the limits recorded by Pischel {Gram. d. Pkt. Spr., §§ 515-17)
for any Prakrit, even for ArdhamagadhI, which goes farther than the
other Prakrits.! gome of these (such as yena . . . tena) are likewise
1 Thi.s use of fist or asi (.Skt. or a.ni) is common Prakrit. In the Saddharma-
pundarika we find not only u--tt or cciiiivalent (as well as n-fti) so used, but also, e.g.,
ahhiit with .subject aham or trnm (SP., 22, 11 and 64, 11 both pro.se) ; and in faet any
third person singular verb may be so used (e.g., ahum . . . akarod, 258, 7). In such
a late text as the Lankavatara Sutra I note (8, 6) aim tdh pamndah mrm ekaikasmin
hi drsyatf (3 .sg. with pi. subject.s). In Pali, atthi (Skt. adi) is used with plural subject
(Geiger, Pali, § 141 ). but that seems to be as far as Pali goes in this direction.
THE PRAKRIT UNDERLYING BUDDHISTIC HA'BRID SANSKRIT 505
found in Pali ; even when this confirmation is lacking, it may reason-
ably be assumed that most of them were inherited from the proto-
canonical Prakrit.’
I find a third indication of the linguistic independence of the hybrid
Sanskrit of the Buddhists in its metrical principles. This subject
requires more extended treatment than I can give to it here ; I am
dealing with it more fully in a paper which I expect to publish shortly
in a volume of studies in honour of Professor Kuppuswami Sastri of
Madras. The metre of such a text as the Saddharmapundarlka is
constructed on principles which in some important respects are quite
different from any found in Vedic or Sanskrit metres, of any period.
I may add that these principles have never been understood, or at
least correctly formulated in print ; and that they were badly misunder-
stood by Kern and Nanjio in their edition, with results which seriously
vitiate the form of the text as printed there. In part, at least, the
same principles reappear in the metres of other Buddhist Sanskrit
texts, such as the Lalitavistara. To mention only one important
feature : the substitution of two short syllables for a long is permitted
ad libitum (with certain definite restrictions in the case of some metres).
This reminds us of the well-known dnja group of metres : but in Sans-
krit the principle is practically limited to that group, which stands quite
apart from other metrical tj-pes ; and even there it is not applied in
the same way. Very scant traces of a similar tendency were detected
by Hopkins {Great Epic, 301) in the epic tristubh ; but they may
perhaps be interpreted differently, and in any case they never amounted
to such a clearly defined metrical licence. Here again I believe that
the hybrid Sanskrit of the Buddhists must be assumed to have inherited
a feature of the protocanonical Prakrit ; for no other origin is easily
conceivable. Moreover the dryd ti'pe, revealing somewhat similar
principles in the one matter just mentioned, is commonly regarded
as of Prakrit origin.
The fourth, and most striking, distinctive feature of this hybrid
^ Similarly, etad abhut {or ahharat. or the like) = Pali etad ahosi„ " this thought
occurred to . . (with genitive). An interesting construction, for which I do not
know a parallel in Pali, but which is rather frequent in hybrid Sanskrit, is md [hniva)
with the optative in the sense of “ isn’t there danger that . . . ? ” SP., 76, 5 (prose)
tat kim manynse sdriputra : md haiva tasya purusasya 7nrsdvddah ■'^ydd . . . “ so what
think you, Sariputra ? i^n't there danger that lying would pertain to that man
(i.e. that he would be guilty of lying) ? ” A useful “ Outline Syntax of Buddhistic
Sanskrit ” has been published by Sukumar Sen in the Journal of the Department of
Letters, University of Calcutta, vol. 17 (1928), It is, however, far from complete;
e.g., it fails to record the md -{- optative construction just mentioned.
VOL. VIII. PARTS 2 AND 3.
33
506
F. EDGERTON
Sanskrit is, of course, the large number of forms which violate Sanskrit
grammar, as to phonology or morphology or both. It is from these,
if at all, that we must hope to discover the grammatical structure and
original location of the protocanonical Prakrit whence they were
taken over.
For this purpose we need first of all a comprehensive grammatical
study of Buddhist hybrid Sanskrit. This is needed, incidentally,
for other purposes, too ; for the philological interpretation of the texts,
and even for the correct editing of them. Most of the existing editions
and translations are quite defective, because they were made without
adequate knowledge of the grammar of the language, not to speak of
its metrical principles and other features. In fact, almost nothing
systematic has been done in this field. The only monograph I know,
aside from Sen's (above, p. 505, n. 1), is Weller’s dissertation, Uher die
Prosa des Lolita Vistara (1915), which limits itself to the grammatically
less important part (the prose) of a single text. Otherwise we have
only the stray observations of individual editors and translators,
which are not only scattered and unsystematic, but often positively
misleading.
The importance and interest of the subject seem to justify, then,
the undertaking of a Grammar and a Dictionary of the hybrid Sanskrit
of the Buddhists ; and this is the task which I have ventured to set
myself, perhaps audaciously, but I hope without minimizing its great
extent and its substantial difficulties. It will require minute textual
study of at least the older and more important literary works and the
relevant inscriptions, and should involve frequent reference to such
Tibetan and Chinese versions as are available. Having been engaged
on it much less than a year, I can speak as yet only on the basis of
very tentative and incomplete results ; in fact, chiefly on the evidence
of the Saddharmapundarika, backed by only casual reading in other
texts.
Unfortunately, as has been intimated, we cannot use the printed
text of SP. uncritically (and this is only too commonly the case with
editions of Buddhist Sanskrit texts). In part the editors may fairly
be blamed for this ; quite often they quote the correct reading in
their critical notes, but introduce a false reading in the text, misled
by erroneous ideas regarding the language or the metrical structure.^
romanized and revised ” edition of SP. by Wogihara and Tsurhida,
^ 0 ^ 0 , 1934 if., of which I have seen the first two parts, corrects some of these errors,
n eaves the majority untouched. It by no means supersedes the Kem-Nanjio
THE PRAKRIT UNDERLYING BUDDHISTIC HYBRID SANSKRIT 507
But in part it was not their fault. Professor Liiders has shown (in
Hoernle, Manuscrijpt Remains, etc., 161) that the Kashgar recension
of SP. contains noticeably more Prakrit forms than the Nepalese version
on which the printed text is chiefly based, though with some reference
to the group of Kashgar MSS. called collectiv’eh" “O’" by the editors.
(Liiders’ observation was anticipated by Kern, Preface to ed., vi.)
Particularly in the prose, the Kashgar fragments show such Prak-
ritisms often enough to suggest that originally the prose of SP. may
have been, like that of the Mahavastu, no less Prakritic than the
verses. It looks as if an attempt had been made to correct ” it
in later times. The verses may have escaped much of this process
because the metre made it more difficult. But they did not escape
it entirely, as Liiders shows (cf. also just below). A complete edition
of the Kashgar recension, if it were possible, would doubtless come
closer to the original form. Yet even it surely suflered some of the
same “ correction ”, since sometimes its readings are less Prakritic
than the Nepalese.
I wish further to emphasize the fact that in the verses of SP.,
initial consonant combinations, which in Prakrit would be simplified,
were always pronounced as single consonants (cf. Kern. Preface to
ed., xi, which understates the facts). For, not only do they fail
to make long a preceding syllable ending in a short vowel ; but even
originally long final vowels, which in this text are regularly shortened
metri causa (but only metri causa, never otherwise !), are shortened
before such combinations, when a short syllable is required, e.g.,
SP., 90, 3 viditvd trdnam (third syllable short ; -tvd for -tea occurs only
metri causa ; here it implies t- for initial tr-). This metrical shortening
proves that the composer pronounced a short syllable, despite the
writing of two initial consonants. Such pronunciation of conjunct
consonants, as if single, is never indicated internally ; that is, it
occurs always, and only, where standard Prakrit phonology would
require or at least permit it. (The beginning of the second element
of a compound is usually treated as initial, though there is some
fluctuation ; this accords perfectly with Prakrit usage.) Conversely,
also, metrical lengthening of a final short vowel occurs before such
combinations ; this necessarily implies the same Prakritic pronunciation,
edition. I cannot refrain from expressing regret, in passing, that the editors saw fit
to compose their footnotes in Chino-Japiincse, a needless hardship for western users
of the book, and peculiarly inconsistent in a work which prints the Sanskrit text in
roman transliteration.
508
F. EDGERTOX —
since if two consonants were pronounced there would be no reason to
lengthen the vowel. So, SP., 27, 15, where all MSS. read vinesijati or
°te ; the former is doubtless to be read, and has metrical lengthening
for °tl before the word prana-, which was, therefore, pronounced
pdna-. Such lengthening is very common metri causa, but never
occurs otherwise.^ Again, in SP., 162, 6, we find a pdda : rayam ca
loJcas ca anugrhitah (or °tdh). The eighth syllable must be long ;
according to the writing, the metre is faulty. Hence the Tokyo
edition emends to anu°. But all MSS. read anu°, and this must be
kept. The word was pronounced amtgg°, as in Pali {anuggahita-)
and Prakrit (anuggaJiia-, °hm-). Uiktwise parigrJiitdh, SP., 89, 8, all
MSS. ; Tokyo edition emends to pari°, because a long syllable is
required, but we must understand parigg°. There are not a few other
metrical indications that originally the language was at least
pronounced (whether written or not) more Prakritically than it is
written in any of our MSS.
It is reasonable to assume with Liiders that where the MSS. differ,
those showing Prakritic forms are more primary than those with
correct Sanskrit forms ; and that the original SP. was “ written in
a language that had far more Prakritisms than either of the two ver-
sions “ (Kashgar and Nepalese). I eannot, however, agree with Pro-
fessor Liiders when he goes on to say that he is “ inclined to believe
that the original was written in a pure Prakrit dialect which was
afterwards gradually put into Sanskrit.” This h}q)othesis makes it
difficult to explain the many correct Sanskrit forms, often quite
foreign to all known Prakrits, which occur side by side with Prakrit
or semi-Prakrit forms, in all manuscripts and frequently guaranteed
by the metre. To mention only a single instance, no Prakrit dialect
has any trace of the Sanskrit perfect, except the isolated dhu (and
dJmmsu), and the like is true of Pali except in artificial Kunstsprache
(see Pischel § 518, Geiger, § 171). But in SP. (including the verses)
perfects, while not very common, are quite familiar, and are used
no more incorrectly than other verb forms. I cannot doubt that they
belong to the original language of our work, which was not a pure
Prakrit but a hybrid dialect, based on a Prakrit, but partially
* Both editions emend to rine.syntt. misunderstanding the matter here treated.
Very rarely do we find a final short vowel before an initial consonant group in a
metrically long syllable. Such cases are not a whit commoner in the JISS. than
before single initial consonants. In all of them some special explanation must be
sought, or emendation resorted to.
THE PKAKRIT UNDERLYING BUDDHISTIC HYBRID SANSKRIT 509
Sanskritized from the start. The extent of this original Sanskriti-
zation is very hard to determine ; certainly it did not go as far as our
editions suggest.
WTiat, now, was the Prakrit, underhung Buddhist hybrid Sanskrit ?
According to Sylvain Levi (cf. p. 502, n. 1 above), p. 511, “ one of the
languages of the land of Magadha.” Liiders (1. c., 162) is more definite ;
on the basis of vocatives plural in -aho from a-stems, which he says
are found ‘‘ only in MagadhI ”, he thinks we may “ assert that the
original text of the Saddharmapundarlka was written, if not in pure
MagadhI, in a mixed Sanskrit which was based on that dialect”.
I cannot agree with so definite a statement as this. The voc. pi.
ending -aho cannot be called exclusively a peculiarity of MagadhI
(see No. 5, below). Our language lacks any trace of some of the most
striking characteristics of MagadhI, such as the substitution of I for r,
and of s for s ; the nom. in e of a-stems was also not characteristic
of it (see No. 11, below). Levi's more cautious formula, “ one of the
languages of Magadha,” may be right, or at least not far wrong.
There is some e\'idence which suggests an eastern origin, at any rate.
But even this can hardly be proved on linguistic e\’idence at present.
Certainly no identification with any known specific Prakrit is possible.
On the contrary, there is evidence which forbids any such identification.
It seems that the language underlying hybrid Sanskrit was different
from any Prakrit known to the later grammarians, at least. It is,
however, possible to find a considerable number of individual points
of agreement with specific Prakrits. And it turns out that nearly
all of them point to either (1) ArdhamagadhI, or (2) Apabhransa.
I have found very few Prakritic features which do not occur in one
or the other of these, and a number which belong to one or both of
them almost or quite exclusively. It is worth emphasizing that the
language was different from Pali in many important ways, while
specific agreements with Pali are very few, minor, and dubious.
The same is quite as true of MagadhI, with which I do not know of
a single exclusive agreement. Indeed. I have failed to find, so far,
any unmistakable evidence of specific agreement with any known
Prakrit except AMg. and Ap. Yet our language also differs from
each of these on important points.
I shall now list briefly the linguistic features of this language
which seem to me to suggest specific agreement with particular
Prakrits, ignoring those which are common to all or most Prakrits.
The following collection, then, contains all the e\'idence now known
510
F. EDGERTON —
to me which could be used in localizing the protocanonical Prakrit
of the Buddhists. It must be remembered that it is chiefly gathered
from a single work, the SP. (to which I refer by page and line of the
Kern-Nanjio edition) ; it therefore makes no claim to completeness
or finality. The prime reference-work for most Prakrits is, of course,
Pischel's grammar ; specific references to it are generally omitted as
unnecessary. For Apabhransa, however, it needs to be supplemented
by the later works of Jacobi (Bhavisatla Kaha, abbreviated Bhav.,
and Sanatliumdracaritam, abbreviated San.), and Alsdorf (Kurnam-
pdlapratibodha, abbreviated Kum.).
1. The nom. and acc. sg. masc. and nt. of a-stems ends very
commonly in either a or u. (The regular Sanskrit forms are also
common ; this may, indeed, generally be taken for granted of all
the forms I shall mention.) Of these, a is common in Ap. and occa-
sional in verses in AMg. and Mg. ; u is recorded by Pischel only for
Ap. and DhakkP (a little-known dialect classed as midway between
Mg. and Ap., and by some grammarians considered a form of Ap.,
though it agrees with Mg. in some important respects such as the
change of r to 1). Certainly a is a phonetic development from as
(ah) or am (am), with phonetic loss of final consonant. Similarly
u in the nom. represents a shortening of o, the common Prakrit ending.
Pischel regards u as phonetically derivable also from am, am. It is
true, at any rate, that ii occurs also for other final om (as well as other
as, o) ; likewise a for other final as and am. E.g. ahu = aho (inter-
jection) 62, 4 and 16; bhuya, 96, 2, and bhuyu, 95, 1, = bhuyas ■,
ahu and hu = aham, 62, 15 and 195, 5, and often (probably also
ha = aham, 195, 4, and 88, 10) ; mahya ~ ynahyam, 86, 8, etc. These
forms are largely regulated by metrical requirements ; they are the
shorts to o, am. T et m also occurs in a metrically indifferent position :
utpannu 177, 9, initial in an anustubh. Were it not for such forms
as ahu = aham (and Ap. tnahu, majjhu = mahyam, etc., Pischel § 351),
one might be tempted to question u from am as a phonetic change,
and regard the acc. forms in u as transferred from the nom., and the
nt. from the masc. For our language seems to have been similar to
Ap. in this, that it tended to make no formal distinction between masc.
1 The «-forms occur also, rery often, in the language of the “ Prakrit Dhammapada ■
of the Dutreuil de Khins MS., edited first by Senart and later by Barua and Mitra.
A systematic linguistic study of this dialect has yet to be made • it has evident
affinities with our dialect, and must certainly be taken into careful’ consideration in
future work on this subject. To identify it with our dialect would be premature to
say the least. ^ ’
r
THE PRAKRIT UNDERLYING BUDDHISTIC HYBRID SANSKRIT 511
and nt. forms (see No. 6, below), nor between nom. and acc. forms.
Namely : —
2. In general, most Prakrit nom. and acc. forms are used inter-
changeably, as in Ap. Since final nasals and anusvara are often dropped,
especially ynetri causa, and final visarga likewise (see the preceding
paragraph), some of these ambiguous forms may be regarded as proper
to either case (i.e. derived by phonetic process from both Sanskrit
forms). However, there are cases where metre, at any rate, cannot
be directly concerned. Thus at the end of a fada, or in an otherwise
metrically indifferent position : sasti. 303, 11, agrabodhi 310, 12,
both nom. Or after a long vowel (as in Ap., Alsdorf, Kum., 58),
trsitdm dJidrani tarpet, 126, 14 (for dhdramm). AMg. and even S.
(Pischel § 379) have noms. in hh, um, regarded by Pischel as the
phonetic equivalents of i, u \ in our text they are rare,^ but cf. loka-
dhdtum nom. 31, 9, and bodhim nom., probably to be read with MSS.
for ed. bodhi in 63, 8. For metrical reasons, the acc. sg. of even fem.
d-stems may be reduced not only to atn (pujam Idrswi, 15, 3, imam
eva dntdm, 61, 11 et passim), but even to a (itna huddhabodhim, 95, 8 ;
carya = carydm, 120, 7, 149, 8, et passim), which also occurs as
nom. of d-stems. Since westerns, as in Pali and Prakrit generally,
often add the thematic vowel a and are declined like u-stems, it follows
that their nom. and acc. forms often coincide, ending in a or u like
genuine n-stems. In the plural much the same state is found. Not
only d (without regard to the nature of the following sound), but
also, and very commonly, a (as in Ap.) occurs as nom. pi. of a-stems ;
the latter, to be sure, apparently only metri causa. Both occur also
as acc. pL, especially a [nirgata . . . ddrakdn, 88, 1), but also d (buddkd
ca bodhim ca prakdsaydmi, 47, 12, for buddhdn). Nay, even the regular
Sanskrit nom. ending dh is used as acc. : magimh, 54, 8, agreeing
with sattvdn ; -purndh acc., 9, 3. The same is found in later texts,
e.g., Lahkavatara Sutra, 6, 5, apsaravargds ca (pratigrhna), where the
puzzled editor suggests emendation. As to i- and w-stems, we find
an astounding variety and confusion in the nom. -acc. pi. forms ;
those actually found resemble AMg. more than any other Prakrit,
and contrast strikingly with the simple state of things in Ap. which
uses i, u for both. The regular Sanskrit nom. in ayas, avas may be
used as acc. : {buddhdn) bahavo, 207, 10, rdtrayo acc. even at the end
of a tristubh-jagafi pdda where rdtri{h) would have done quite as well
^ There are clear cases in the LalitaTistara, e.g. 49, 16 (Lefmann), na cdsti tTptirh
(aU MSS.).
512
F. EDGERTON —
metrically, 91, 3, and even in prose dunduhhayas as acc. 69, 11.
Besides, we find x (and metrically i), inas, is (even as nom. masc. !),
yas as acc. (fem.), and in the fern, lyas, iyo (before a surd, 86, 1), iyd
(before ca, 237, 3) and iyas as nom. or acc. indifferently. As in Ap.,
we thus find the language far advanced on the road to a declension
containing only two forms in each nmnber, a nom.-acc. and an obUque
case, though the forms differ from those actually found in Ap. (The
obhque cases, especially of the fem., are much confused in Prakrit
generally.) It is as if, in this respect, we were dealing with an immediate
precursor of a modern vernacular. The same confusion occurs in
pronominal forms : yuyam as acc., 198, 1.
3. Very common are neuter nom.-acc. pi. of a-stems in d (sporadic
in various dialects, but especially AMg.) and a (usually metri causa ?
regular in Ap.) : bald, 62, 2 ; dmtrinsatilahsana mahya bhrastd (v. 1.
°pi) 62, 1. Though these have been interpreted as inheritances from
Vedic, it is quite as likely that they are merely taken over from the
masc., where they are very common (as we just saw) for Sanskrit
as, an ; cf. No. 6, below.
4. Besides the general Prakrit ending «, we find u in the nom.-acc.
pi. of M-stems. This is not limited to syllables where the metre
requires a short : bahu me dhartna bhdsitdh, 255, 7, in anustubh metre.
This short u is not recorded by Pischel ; it seems to be characteristic
of Ap. (Alsdorf, Kim., 59).
5. Voc. pi. of a-stems in dho. Quoted by Liiders (l.c., see above)
as exclusively 3Ig. But Ap. also has aho, ahu. The forms actually
recorded by Jacobi and Alsdorf seem to show only short a in the penult,
but this is probably a mere accident. They are not numerous in any
case ; and it is an established principle of Ap. that stem-vowels in
penultimate syllables may be either short or long (Jacobi, Bhav.. 28*,
San., 1, 9, 12 ; Alsdorf, Kum., 55). Even the original d of feminine
stems is shortened frequently (usually, according to Jacobi, Bhav.,
I.C.). The voc. pi. certainly contained d(ho) originally, and it seems
to me that our SP. form may much more plausibly be regarded as
a link with Ap. than with Mg., since there is no other special agreement
with Mg. The ending is not common, yet is sufficiently well authenti-
cated ; e.g. in hdaputrdho, 253, 1. and 255, 11. All MSS. apparently
have arnaresvardho in Lalitavistara (Lefraann), 47, 5. It is not recorded
in AMg.
6. The pronoun so, properly masc., is also used as nt. nom. and
acc. ; 60 (= tad) eva vicintayantah “ pondering this same thing ” 62, 7.
THE PRAKRIT UNDERLYING BUDDHISTIC HYBRID SANSKRIT 513
So Ap., and (in the forms se, se) AMg. and Mg. (Pischel § 423). Jacobi
and Alsdorf do not quote Ap. so, su as nt., but San., 501, 3, has su
hhavanu acc. nt., and both Jacobi and Alsdorf recognize ehu, ihu ( = em)
as nt. This is probably to be regarded as part of the breakdown of
the Sanskrit system of grammatical gender which characterizes Ap.
and AMg. While some change of gender occurs in the other
Prakrits and even in Sanskrit, it is these two dialects, and especially
Ap., which carry it farthest ; indeed to a point where, as Jacobi says
(BJiav., 31* f.), it is hardly possible to distinguish any longer between
masc. and nt. in Ap. ; and even the feminine is involved in the con-
fusion. The verses of SP. approach this state. Very many nouns
vary in gender, or at least show forms (in their own declension or that
of modifiers) that were originally characteristic of different genders,
and that, too, in close juxtaposition with one another. So in 87, 7 ff.
the noun ydna is modified by both masc. and nt. adjectives and pro-
nouns, in the same context. Masc. cati'am(h) and nt. catvdri both
go with the feni. noun parsdh, 9, 1, 294, 11 (but parsa catasra[h],
fern., 25, 1) ; fern, annttaram with the nt. noun jHd«n(>/() 10, 5 (perhaps
influenced by thought of the fem. synon}’m bodhi 1) ; nt. yavanti
with the masc. form sattvdh, 9, 5, etc. The feeling for the distinctive
generic force of the Sanskrit noun endings, and especially for the
difference between masculine and neuter, was evidently very feeble.
7. Final e, o very commonly become i, u when the metre requires
a short. (In such cases e is occasionally, but rarely, retained in the
YTiting ; I have noted only a few cases of te, 85, 12 and 13 ; 131, 4 ;
152, 11 ; curiously mi seems to be regularly written for me in such
cases.) This reminds us especially of Ap., but it occurs also in verses
in AMg. and other dialects (Pischel § 85). Since me is not an Ap.
form, and mi = me is very common in SP., we may possibly regard
this as a link to AMg. rather than Ap. In the loc. sg. of a-stems,
i for e is specially frequent (so also Ap.).
8. Ap. is peculiar among Prakrits in confusing the instr. and loc.
pi. (Jacobi, San., 11). AVe may see the influence of such a dialect
in 85, 1, vilokayanti gavdksa (v.l. °kse) iillokanakehi ‘‘they look out
at window(s) and loop-holes The parallel yavdk.se indicates that
Iillokanakehi is felt as loc.
9. In 67, 11 occurs the nom. sg. form tuhaih = team. Pischel
records it only for Dhakki ; it is elsewhere attributed to Eastern
Apabhransa, which perhaps means about the same thing (Jacobi,
San., XXV ; cf. Alsdorf, Kum., 59). Another nom. tuva occurs. 93, 9 ; it
514
F. EDGEKTOIf —
stands of course for tuvarii, with metrical loss of anusvara. The form
t'uvam is Yedic (by Sievers’ Law, cf. Edgerton, Language, 10, 235 ff.)
and occurs also in Pali (under conditions no longer regulated by
Sievers’ Law). It seems, according to Pischel, not to be recorded in
Prakrit literature. Since, however, it is quoted by the Prakrit
grammarians, no special significance should probably be attributed
to this seeming agreement between our dialect and Pah.
10. The same holds good, I think, of the only other formal
agreement with Pali which I have noted. Pali has oblique case
forms of a-stems in a, besides aya, ayd. As Geiger (§ 81, 1) says,
this is evidently a contraction of the other forms, or of the Prakritic
da (or de). At least one such form occurs in SP. : disd, loc., 191, 5.
Since contraction of vowels after loss of an intervening consonant is
fairly common in Prakrit generally, I am not inclined to attribute
much significance to this agreement with Pali, though the form seems
not to be recorded in Prakrit.^
11. Very rare is the AMg. Mg. nom. sg. masc. a-stem ending e,
for normal Prakrit o (note that even Dhakki has o). Clearly o (whence
u, Nos. 1 and 7 above) was the regular ending in the Prakrit under-
lying our dialect. I have not found e in any substantive.^ Perhaps
the only clear case of it is uttare, 313, 8, which can only be nom.
sg. masc. : no v.l. is recorded. Less certain is ke-cit, 115, 2, where the
Kashgar reading is kimcit ; furthermore, it is not impossible that
kecit is meant as a plural (the noun is hhogu according to the reading
adopted in the edition, and it is doubtful whether u can pass as a
nom. pi. ending of an a-stem ; but there is a v.l. bhdga, which might
easily be nom. pi.).
12. We have referred above to the extensive use of 3 sg. verb
forms with 3 pi., and also 1 and 2 sg. subjects. This seems to be
characteristic of AMg. (Pischel §§ 516-18), which goes much farther than
any other known Prakrit (for a possible trace in Ap. see Alsdorf,
Kum., 65). AMg. also uses 3 pi. forms in the same way (l.c.). In
SP., at least once, a 2 sg. form abhuh is used with 3 pi. subject (176, 12 ;
well attested in both recensions ; only one Nepalese MS. abhut).
^ I reserve for another occasion a fuller discussion of the oblique cases of fem.
nouns, merely observing that the usual endings (when not regular Sanskrit) in SP. are,
for all oblique cases, aya, iya, uya. The first of these agrees precisely with Pali, but
Prakrit (ae, aa) is not far removed. In other Buddhist works we find aye (abhdye,
instr., Lalitavistara, 122, 20, Lefmann), lye {kmniiye, ibid., 162, 3), etc.
® But note Lalitavistara (Lefmann), 74, 4, bodhisattva brahmakalpasarhnibhe
(nom. sg. ; no v.l. recorded). This is the only case thus far noted in LV.
THi: PRAKKIT UNDERLYING HUDDHISTIC HYBRID SANSKRIT 515
I have not noticed a 3 pi. with sg. subject, but in 108, 17 (prose) I
believe we must read abhuvan, with most Nepalese MSS. for ed. abhuma
(subject vayam ; Kashgar MSS. dsit). All sorts of 3 sg. forms are used
indiscriminately with subjects of all persons and numbers ; they
include optatives, perfects, etc.
13. AAIg. has verb forms in e which look like optatives but are
used as past indicatives (Pischel § 466, end), and in general, as
Pischel there shows, AMg. reveals a strange confusion between optative
and aorist forms. Our text seems to have the same phenomenon.
In 190, 7, sprse can only be past indie, in meaning (in describing a
past Buddha’s attainment of enlightenment, sprse sa bodhim) ; usually
such a form is optative (=sprset) but that is quite impossible here.
Conversely, forms in I occur, which look like aorists (Skt. -It), but
seem to be interpretable only as optatives ; e.g., 291, 12, sarvesu
maitribala so hi darsayJ “ he shall show the power of kindness to all
beings”. Metre cannot be concerned here, since it occurs at the end
of a pada. (Similarly 295, 2, 4, 7, 8.) The explanation is obscure ;
probably it is connected with formal, phonetic confusion between i
for e{t) in the opt. and I for i{t) in the aor. In any case we have here
another, and a rather striking, agreement with AMg.
14. Fairly common is the 2 sg. imperative ending aki. It seems,
according to Pischel (§ 468), to be specially characteristic of AAIg.
and (in the form ahi) Ap., though it occurs occasionally in other
Prakrits. It is also known to Pali.
15. Quite frequent are presents of the tjqie kurva-ti, from the
root kr. They are found only in AJVIg. and (evidently under the
influence of this canonical language of the Jains) in Jaina Maharastri and
Jaina SaurasenI ; not in Ap. Since Pali also has kubbati, but evidently
as a borrowed form, not native to the dialect (it is used chiefly in
gathas), we may conclude that it was peculiar to the protocanonical
Prakrit and to AMg. (of course in the form kuwa-i).
16. Another striking agreement with AMg. appears in the gerunds
in -ydna = AMg. -ydna{m). peculiar to that dialect (Pischel § 592),
e.g. sruniydna, 61, 9, etc. ; about a dozen instances have been found
in SP.
17. On the other hand, gerund-forms in i (and I) point rather to
Ap., where alone i is recognized as a gerund-ending. Pischel §594
explains it as for Prakritic -ia with loss of final a (query ; rather
directly from -ya by “ samprasarana ” ?). Jacobi does not recognize
it as occurring in his Ap. texts, but there is at least one case in San.,
suni, 445, 5 (“ having heard that the sun, the friend of the world, had
516 THE PRAKRIT UNDERLYING BLIDDHISTIC HYBRID SANSKRIT
gone to rest ’') ; and tkree forms occur in Kum. (Alsdorf, 63). In SP.
the ambiguity of the ending makes the interpretation often doubtful ;
for the same ending occurs in the opt. and aor. indie, (above, No. 13),
and sometimes even a noun form is conceivable (nom.-acc. sg. or
pi. of i- or iw-stem). But there are cases where it seems to me that
any other interpretation than as gerund is implausible : abhyohiri,
325, 4 ; upasarhhraml, 11, 11 ; karapayi, 152, 5 (to be rendered “ and
after causing stupas to be made for them when they have entered
nirvana, he will honour them ", etc.) ; upasaniTcraml , 191, 1 (note that
this clause stands between two precisely parallel clauses, in both of
which the verb form is an unmistakable gerund, viditva and abhyoki-
riydna ) ; na uttarl prarthayi ndpi cintayi, 213, 10, (for we were satisfied
with mere nirvana,) not asking for, nor even thinking of, anything
further." Perhaps also abhyokiri, 228, 15 (which, however, might be
considered 3 sg. opt. with Burnouf and Kern).
18. “ Short vowels, internal as well as final, are very commonly
lengthened for purely metrical reasons, especially in AMg. and Ap.’
(Pischel, § 73). And further : “ In Ap. verses, long and short vowels
interchange according to the needs of meter and rhyme (ib. § 100).
In our dialect it is no exaggeration to say that any vowel may be
lengthened or shortened to fit the metre. It is mostly final vowels
which are treated so cavalierly ; they are lengthened and shortened
without the slightest compunction, and so commonly that examples
need not be quoted. But also internal vowels : adhyesami, 1 sg., for
°ami, 38, 2 ; khudrdka for kmidrakdh, 127, 3 ; andbhibh uh for 128, 4.
I regard this as another link with AMg. and more especially with
Ap. ; no other Prakrit goes so far as these two. It should be added
that the regular Prakrit “ law of morae " applies here too. (The
best statement I know is in Geiger, Pali, § 5f. : double, i.e. long,
consonant, and also short nasalized vowel, may interchange with long
vowel at any time, without regard to etiunological origin.) Hence,
instead of metrical lengthening of a final short vowel, it may be
nasalized, or the initial consonant of the next word may be doubled.
So sddJmm { = sddhu) ti yhosam, 55, 12 (in the very next line occurs the
equivalent sadhii ) ; dasa-ddimsu, 32, 14 and often, also dasasu-ddisdsu,
55, 11, etc. For further details see my article soon to appear in the
volume in honour of Professor Kuppuswami Sastri.
I think this evidence is sufficient to indicate that the protocanonical
Prakrit, on which Buddhist hybrid Sanskrit was based, was a dialect
closely related to both Ardhamagadhiand Apabhran&, but not identical
with either.
Alphabets and Phonology in India and Burma
By J. R. Firth
LlOR 300 years after Vasco da Gama touched Calicut generations of
traders, merchants, missionaries, soldiers, and other emissaries
from at least five different nations of Europe took their turn in India,
pursuing their interests at a respectful distance, making no obtrusive
efforts to scrape acquaintance with Sanskrit culture,^ Such advances
were socially difhcult, and would not have been welcomed, Jlore-
over, our early associations were with Dra vidian India, and very few
cultured Brahmins sought membership of Christian Churches,"
As late as 1771 Amaduzzi, the head of the Typographia Sacrae
Congregationis de Propaganda Fide, writing of the Alphabetum
Bratnmhanieion seu Indostanum Universitafis Kasi, remarks : “ Cui
etiam Historiae, Fabula, Scientiae, ceteraque pvcr-qeLa commendantur
ne ceteris de plebe, ac peregrinis quinetiam arcana huiusmodi
patere possint, Quare Idioraa hoc ab ipsis Samscrit
appellatur ” . , , “ Eadem Lingua Samscritica, seu litterali Bramm-
hanica pro sacris, et arcanis rebus singulae hae gentes religiose,
constanterque utuntur,”
The Capuchin missionaries,® upon whose work the Alphabetum
is chiefly based, report : ‘‘ Brammhanes tamen, ut iam innuimus,
maximo studio, turn zelo servandi Religionis arcana, turn metu
punitionis subeundae, non solum aUenigenis, sed terrigenis etiam, qui
de eorum tribubus non sunt geniti, abscondere solent huius Alphabeti
institutiones.” The knowledge of the " Bedpurana ’’ is the secret of
the few “ ceteris autem perpetuis in tenebris delitescat ", And so it
was with the excellent Capuchin friar, Beligatti, as with so many
generations of Europeans in India, Besides, had not the worthy
^ The Italian students of Sanskrit, Sassetti (1581-8) and de Nobili (d, 1656)
were the exceptions proving the rule,
^ The Tamil teacher and interpreter employed by Ziegenbalg in 1706-7 was,
wo are told, expelled from Tranquebar and subsequently kept in irons in a Tanjore
prison, accused “ d’ avoir trahi la Religion, et d’en avoir revele les Mysteres le plus
secrets aux deux Mi.ssionaires de Tranquebar ! ” La Croze, llisloire du Christianisme
des Indes, tome ii, p. 391.
* Especially Cassiano Beligatti di Mar-erata, also joint author of the Alphabetum
Tangufanum sive Tibetanuin (1773). Worked in the Tibet-Xepal Mission. In Lhasa
1741-2, then twelve years in Nepal and occasionally in Patna. Lied in Macerata 178.5.
518
J. R. FIRTH
Magister Balgobinda of Patna told him that he himself found
difficulties with Sanskrit every day and there was no end to learning
it ? And so, like many both before and after him to this day, he says :
“ At cum nobis concessum non fuerit talia penetrasse secreta, ut
aliis ad ulteriora perscrutanda planam viam panderemus.”
Alphabets with all their implicit phonetics, phonology, and
grammar, have a background of at least 2,000 years of history in India,
and to this day they remain the totems of the peoples, marks of
brotherhood, and against the stranger graven shibboleths. Alphabets
divide and rule. We English, following the opinion of Lord Macaulay,
pressed our ABC and the rest of our literary arcana on our Indian
fellow subjects. The interesting thing is, however, that the passwords
of the English “ governing voice ” continue to serve in phonetic
safety the close fellowship of the ruling caste.
Contact with the vernacular languages was different. Even in the
earliest days, of course, Europeans on arrival in India had, as we
say, “ to learn the language,” and superficial knowledge of certain
vernacular languages necessarily started with the first systematic
relations between Europe and India. More scholarly acquaintance was
especially necessary for the great missions, and so we find that a
study of the vernacular languages long antedates what we are pleased
to call the discovery of Sanskrit, following Sir William Jones’s epoch-
making address in 1786.
In spite of the early neglect and ignorance of Sanskrit, in spite of
Lord Macaulay's appalling judgment, Eiu’opean scholars and especially
Englishmen have during the last 150 years served Indian scholarship
well, and none better than Sir George Grierson. Just as Macaulay's
minute, in establishing the use of Enghsh in India, inaugurated the
biggest Imperial language and culture imdertaking the world has
ever seen, so this monumental linguistic survey of a vast sub-
continent is the biggest thing of its kind in history. Both in devoted
labour of direction and in the piety of its collaborators it holds perhaps
the highest place in the long history of such work in India.
This is a fitting occasion to recall the work of the earliest students
of Indian languages, Portuguese, Italians, Dutch, Danes, Germans,
Frenchmen — especially of the missionaries. Catholic and Protestant,
and also the pioneer publications of the Press of the Sacra Congregatio
de Propaganda Fide in Rome and the Tranquebar Mission.^ In
* Most of these earlier works are mentioned in the Linguistic Survey of India.
See vol. iv, pp. 302, 350 ; vol. v, p. 18 ; vol. ix, pp. 6, 7, etc.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 519
1771 it was probably justly claimed that ; “ Ceterum nullus forte
locus Bramiuhanicis, et aliis quinetiam exoticarum linguarum
Codicibus magis abimdat, quam Bibliotheca Collegii Urbani de
Propaganda Fide . .
K we suspend for a moment all theological notions of linguistic
unity, the noises of the human race are indeed a chattering Babel,
a confusion of tongues. Such abounding diversity is at once a challenge
to those minds which seek ordered simplicity in the world, and at the
same time a collector’s paradise. There will always be those who seek
an underlying unity, and both theology and historical philology have
immensely strengthened this way of regarding the languages of the
world. But, in spite of the “ philological revolution ”, the traditions
of plain description and the enthusiasm of the field collector have con-
tinued unbroken, from Gesner’s Mithridates in 1555 to the collections
of the International Phonetic Association and the recent Internationale
Arbeitsgemeinschaft fiir Phonologic. In 1592 Hieron}Tnus Megiser
of Stuttgart printed Specimens of Forty Languages, increased to fifty
in the second edition of 1603. But it was not until the eighteenth
century that the systematic collection of material was undertaken
in earnest. Leibniz stimulated his many correspondents and interested
Peter the Great. And it was. in fact, in a letter ^ of Theophilus Siegfried
Bayer, one of the founders of the Imperial Academy at St. Petersburg,
that the first words intended for Hindustani were published in Europe.
In another (1729) we learn of records of the Sanskrit alphabet.
The first really comprehensive compilation was the Orientalisch- und
Ocddentalischer Sprachmeister, by Johann Friedrich Fritz imd Benjamin
Schulze, published in Leipzig in 1748. It presented 200 translations
of the Lord’s Prayer and 100 alphabets, including the Bengali, and the
Modi alphabet for Marathi, Gujarati, and Tamil, Telugu, and Canarese.
It was the first collection of Alphabeta in which Indian vernacular
words were printed in their own character in movable type. From the
phonetic point of view it falls far short of the later Alphabeta of the
Press of the Sacra Congregatio de Propaganda Fide. No notice is
taken, for instance, of cerebral or retroflex consonants. But it held
the field till 1771, when the Alphabetum Brammhanicum was
published, marking a new epoch in Indian studies.
The linguistic employment of the parable of the Prodigal Son and
the fable of the North Wind and the Sun, follow directly in the
tradition of the Sprachmeister.
^ 1st June, 1726.
'V
520
J. R. FIRTH —
At this point perhaps we should notice the very early work of
individual students of Indian languages, and mention the career of
Maturin Veyssiere La Croze.
The first real account of Hindostani was not published till 1743,
though the work was done much earlier by J. J. Ketelaar, a Dutch
envoy to Bahadur Shah, who was in Lahore in 1711 and moved to
Delhi with the Emperor. Later he became Dutch director of trade at
Surat. 1 There are also several early Dutch accounts of Tamil, but the
first systematic grammar, published in 1716, was the work of
Bartholomew Ziegenbalg, a German member of the Danish Mission at
Tranquebar, “ admire des Indiens pour la connaissance et I’usage de
leur Langue." ^
La Croze gives an amusing account of how Ziegenbalg and
Plutschau learnt “ Damul ”. “ Comme la langue Portugaise est depuis
plus de deux siecles fort commune dans les Indes, ils jugerent a propos
de I’apprendre la premiere. . . With the aid of Portuguese they
applied themselves to the study of Tamil, but found the books “ ecrits
en cette Langue sur des feuilles de Palmier. C’etait la tout le secours
sur lequel ils pouvaient compter, y comprenant la vive voix des gens
du pays, qui n'ont ni grammaire ni dictionnaire, ni aucun art qui
facilite I’intelligence de leur Langue They made little progress
so they engaged a Tamil schoolmaster, who brought his school of
small children with him and the two missionaries began “ a ecrire
comme eux avec les doigts sur le sable les lettres Malabares, et a les
joindre selon que le Maitre d’Ecole les dictoit ”. Unfortunately the
master knew no Portuguese, so they were left in the dark as to the
meanings of most of the words they learned to write and pronounce.
But eventually they found a Tamil who spoke Portuguese, Danish,
Dutch, and German ! " Get homme leur fut d'un grand secours, aussi
bien qu un petit abbrege de la Langue Malabare qui leur tomba entre
1 The Alphabetum Brammhamcum mentions a “ MSS. Lexicon, Linguae
Indo&tanicoe in Bibliotheca Collegii Urbani de Propaganda Fide, quod Auctorem
habet Franciscum M. Turonensera ex Capuccinonini Familia, qui ipsura in Suratensi
Missione, quae eidem erat concredita, concinnavit. ac dein dono dedit Sacrae huic
nostrae Congregationi a.d. Ill Nonas Quinctiles anni cniocciv (1704).” This MSS.
is said to contain 489 pages in pt. i and 423 in pt. ii, giving Latin words in alphabetical
order in the first column, “ altera Indostanicas Xagaricis apicibus exaratas.” On the
opposite page the Latin words are said to be written and explained in French in the
first column and in the second, the “ voces Indostanas ” are, “ quantum potis est,”
also written and explained in French.
“ b. 1683, d. 1719. For further details of hi.s life and work see La Croze, Histoire
du C hristianisme des hides, a La Haye aux depens de la Compagnie, 3rd edition, 1758,
vol. ii, pp. 384 ff.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 521
les mains, et qui etoit de la composition d’un IVIissionaire Portugais . . . .
Ils se formerent en pen de temps a la prononciation qui est extreme -
ment difficile.” I suspect that in other parts of India and Burma also
the works of earlier missionaries, even manuscript notes “ fell into
the hands ” of those who eventually wrote the first real grammars
and dictionaries, and established traditions.
In 1716, after completing his grammar on the voyage to Europe,
Ziegenbalg preached before the King of Denmark at the siege of
Stralsund, and afterwards one of his Indian converts had the honour
of being presented to His Majesty. He was received by the King and
the Prince of Wales during his xdsit to England, where he had received
the liberal support of tbe Archbishop of Canterbury and the S.P.G.
since 1709.^ The S.P.C.K. had given him a printing press, which had
been set up in Tranquebar in 1711.
His phonetic observations follow the Tamil syllabary, and though
sound enough in their way, are not especially interesting. The D in
Gmmmatica Damulica is good German and not really bad Tamil.
He noticed the palatal nasal which he transcribed yn and the pre-
palatal affricate for which he used five roman letters, ytsch. He
counted eighteen consonants, five long vowels and five short, and two
diphthongs.^ Like many others who followed, even after the publica-
tion of the Alphabetum Malabaricum in 1772, he wrongly described
Tamil as the Malabar language.® We even find Pope * saying
Malayalam “ seems to be but a corrupt Tamil ”. The Jesuit Beschi
arrived in India about 1700 and produced a new Tamil Grammar (1728-
1739), which seems to have been used by most of his successors. He
had the reputation of being a good Telugu and Sanskrit scholar as
well. He died about 1746.
One of the most interesting personalities in the history of Oriental
scholarship during the early years of the eighteenth century is Maturin
Veyssiere La Croze. He was born at Nantes in 1661 and was educated
by Benedictines, taking a great interest in the writings of the early
Fathers. In 1682 he went to Paris and soon became known on
account of his independence of character and unorthodox views.
In 1696 he had to leave France, and went to Berlin, following the
* Of this support I.,a Croze, w ho was an admirer of England, remarks " Rien n’est
plus edifiant que la charite de la Nation Angloise, qui se signala en cette ocr,asion ”,
loc. cit., 2nd ed., vol. ii, p. 416.
- Cf. my “ Short Outline of Tamil Pronunciation ” in Ardeti'n Qiammar, p. vi.
® See Alphabetum Grandonico-Malabarieum, 1772, p. xxi.
* .4 Handbook of the Tamil Language, oth edition, 189.5, p. 2.
VOL. VIII. PARTS 2 AND 3.
34
522
J. R. FIRTH —
example of many exiled French Protestants, who had been welcomed
there by " The Great Elector (1640-1688). In 1697 he became ■
liibrarian and Antiquary to Frederick. Elector of Brandenburg,
afterwards in 1701 the first King of Prussia. In 1725 he was
ffiven the chair of philosophy in the French College in Berlin,
and he died there in 1739. He wrote histories of Christianity in
India, and in Ethiopia, and from Berlin carried on a voluminous
correspondence with most of the linguists of his time, including Leibniz.
Bayer and Ziegenbalg, mentioned above, and among many others with
John Chamberlain and David Wilkins in England. After his death this
correspondence was published in Leipzig in 1742 as Thesauri EpistoUci
La Croziani. This collection may be regarded as the focus and index
of most of the Oriental linguistic work of the early eighteenth century.
Though he was no friend of the Roman Church his letters are constantly
quoted in the publications of the Press of the Sacra Congregatio in
the last thirty years of the century from the Alphahetxim
Brammkanicum of 1771 to the revdsed edition of the Alphahetum
Barmanorum of 1787.
In view of the discovery of the Tell el Duweir Vase in 1933 and the
still more recent researches of Mr. Starkey at Lachish in Palestine,
which have furnished the missing link in the evolution of the Semitic
and other alphabets from Ancient Egyptian, it is interesting to quote
the La Croze letters. In his letters to La Croze, Ziegenbalg expressed
the opinion that all the alphabets used on the Malabar and
Coromandel coa.sts, in Ceylon, and other parts of India were derived
from the Sanskrit alphabet used by the Brahmans. La Croze him-
self in letters to Bayer and John Chamberlain suggested a common
origin of the Phoenician, Syrian, Arabic, Persian, and Brahman
alphabets, and also hazarded a guess that thev all derived from
Egyptian hieratics and hieroglyphics.*
Giovanni Cristofano Amaduzzi, who presided over the Press
of the Sacra Congregatio, in his preface to the Alphahetum
Brammhanicum of 1771, was well acquainted with these views, and
expressing some doubt continues ; ‘‘ Nisi etiam dicere velimus Indo-
stanum Alphahetum profluxisse ab alio antiquiore Brammhanico
^ See tom. i, letter xiii, p. 16 ; tom. iii, letter ix, pp. 22, 23 ; letter xlii, p. 85 ;
and letter cccxix, pp. 381 et seq. “ J’ai entre les mains les Alphabets Tartares de
Tangut, et des Manchous, ceux de Bengale, de Ceylan, de Malabar, de Siam, etc., en
partie manuscrits, et en partie iraprimes ; et je n’ai point eu de peine a me convaincre,
que tous ces alphabets n’ont eu autrefois qu’une seule et meme origine.” La Croze,
loc. cit., tome ii, p. 246. See also p. 353.
ALPHABETS AHD PHONOLOGY IN INDIA AND BURMA 523
non admodum absimili ; siquidem, teste Cassiano ^ nostro, extant
nunc in Indostanicis Regionibus antiqui Codices apicibus quibusdam
exarati, quos et ipsi peritiores Bmmmlianes se ignorare ingenue fatentur,
dum interim apud ipsos traditio est, neque eorum maiores, a quibus
eos acceperant, huiiismodi litterarum, et nexuuni praesertim. qui
frequentes sunt, potestatem calluisse.”
The various Alphabeta of the Press of the Sacra Congregatio de
Propaganda Pide are abiding testimony to the work of the Capuchin
Friars and other regular clergy working in India and Burma.
Some of them contain phonetic and phonological observations quite
similar to more recent ones which have brought faint thrills of
discovery to observers even in our own time.
The Alphabetitm Brammhanicum of 1771 presents notes on three
alphabets, and below are a few which are of general interest, showing
that these early observations had some phonetic and even phonological
value — largely because the alphabet, though interpreted to Europeans
in Latin or Italian terms, was presented also from the Indian point
of view.
There is, for example, a clear separation of the unaspirated from
the aspirated consonants, which though transcribed as at present by
means of digraphs, ph, bh, etc., are classed as simple and not two
sounds. Aspiration would be at once apparent as the observer notes :
“ Quod apud Latinos non in usu habetur.” The mention of “ in
interiore gutture is especially suggestive of some glottal correlation ;
thus : “ Alios vero obscure in interiori gutture formant, et voce
tenui ac quasi dimidiata profcrimt. Alios quadam vi, et aliqua
aspiratione exhalant.”
Two other prominent phonological characteristics are noted, the
cerebral or retrofle.x consonants and the use of nasalization. “ Alios
insuper medio palato, scilicet ad palatum ipsum linguam inflectentes
emittunt. Alios tandem nariuin ministerio pronunciant.”
The dental t and d are, of course, at once recognized : —
“ a nostro t non differt ”
“ d, est nostrum d Latinum.’’
Of the retroflex “ Aliter quam per t. haec a nobis Latinis explicari
nequit, nec describi potest ; quamvis longe sit diversa, eiusque
pronunciationem assequi necessarium sit. Profertur lingua paullulum
‘ Cassiano Beligatti, mainly responsible for the Alphabetum.
- Alph. Bramtnh., pp. xii, xiii.
524
J. E. FIRTH —
inversa, et palatum leniter percutiente, quo blese pronunciatur.”
"Leniter percutiente” of 1771 shows much more feeling than “ the
tip must he pressed firmly against the highest part of the roof to form
the obstruction and kept in this position for the greater part stop ”
[sic\ of 1915.1
Eemembering Behgatti’s source of information and probable
acquaintance with the dialects of Bihar as well as Nepali, the
following note on retroflex d is interesting. It is given as “ da vel
ra", with the remark: “ duplicem huius litterae pronunciationem
habes ; nunc enim, ut d blesum, nunc ut r itidem blesum, sed palatum
s imil iter leniter percutiendo pronunciatur. Initio quidem dictionis
semper ut d blesum pronunciabis, sed in medio, et in fine certa non
habetur regula, modo da, modo ra dices.”
On the ai&icate ^ (c), transcribed cm, the note is discriminating.
Unlike many less scrupulous writers of later centuries, he rejects
the Italian da. ” Neque haec ulli ex nostris litteris rite potest
assimilari.” Of the voiced correlate of this, transcribed gia, the note
runs : “ Nostro gi et z simul arridet haec littera.”
MTiat is said of ^ {Ba) is also interesting : Latinum 6, de quo
adnotes velim quod sicut in nostra Europa apud varias nationes b in
n, and v consonans in 6 immutatur, et unum pro alio usurpatur,
ita apud Indos invenies dicentes vap-h pro bap-h, {vapor) et vavo
pro bavo.” It is pointed out, however, that “ bha non convertitur
in V
Of m there is the curious note ; “ quod debet aperto ore pronunciari
. . . et obtuso effertur ore.”
He distinguishes long and short vowels of the three types a, i,
and M, and on the basis of the alphabet, groups what he transcribes
as e, ei, and o, au, as similar pairs.
Of “ Bisarka ” he says : “ nullum proprium habet sonum, sed
tantum indicio est litteram, cui iunctum est Bisarka, proferre debere
fortiter, ac si trahereiur e peclore, sono tamen rninime in longum
protract o.”
In 1772 the Press published its Alphabetuni Grandonico-
Malnbaricum sive Samscntdonicum, largely the work of “ Clemens
Peanius Alexandrinus ”, a Discalced Carmelite of the Verapoly
Mission in Cochin. His title was intended to distinguish
the literary alphabet from what he calls Malean-Tamuza or
* Noel-Armfield (on retroflex consonants) in Gev^ral Fhonelics, pp. 9S-100.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA
525
Malabarico-TamuHcam — but be was under no misapprehension as
to the language represented, “ quae proprie Maleaima vocatur, . . .
lingua nova et incognita.’’ He protested against the confusion of
Malayalam with Tamil by Ziegenbalg, La Croze, and others, though
apparently without effect on Pope, who described it as a sort of corrupt
Tamil in the fifth edition of his Tamil grammar, published in 1895.
“ Quasi idem prorsus esset Idioma Malabaricum atque Tamulicum :
quo sane nihil absurdius hac in re comminisci potest. Asserere enim
Tamulicam Linguam Malabaricam esse, aut Malabaricam esse
Taniulicam, idem prorsus esset, ac si Gallicam diceres, et Italicam
Linguam invicem non differre. Licet autem utriusque Linguae,
Malabaricae, et Tamulicae radices communes sint.”
Peanius recognizes five long and five short vowels and two
diphthongs. He presents various tj’pes of syllable such as those
“ quibus copulatur ja, seujota’’, which we may describe as yotized ”.
Then there are combinations with r "I" Latinonim, v, and
also the characteristic doubling of consonants. Of a common final
“ I ” he ivrites : “ linguae injlexione pronuncintur." Of another. M-hich
we sometimes think of as an r sound, he says it is rather like “ r final!
Latinorum ; sed pronunciatur cum aliquo sibilo, clause ferme ore,
ac retrorsum attracta lingua ; idem tamen sonat in medio, ac in fine
dictionis.’’ Not at all a bad description of one of the most difficult
sounds of Tamil and Malayalam.
The homorganic nasal on-glides to the voiced stops are noticed,
which we may represent by 'k, 'p, 't, etc., in the Indian way.
The author notices several characteristic features of the language,
especially the contrast between the lax pronunciation of single p,
t, k, as 6, d, g} in intervocalic position, and the energetically articulated
voiceless stops usually termed “ double ” pp, tt, kk.
“ Ka in principio dictionis aequivalet nostro h ; in medio autem
pronunciatur ut ga.
“ klca est idem duplex maiori vi prolatum, . . . profertur cum aliquo
conatu . . . maiori tamen vi, quod in litteris duplicibus semper est
observandum."
A second prominent characteristic of Malayalam is the
palatalization of consonants. Peanius appears to have noticed this in
distinguishing the two r sounds (in addition of course, to the retracted
^ My own obsen'ations of the pronunciation of a native of Trivandrum made some
years ago seemed to indicate that the lax intervocalic consonants here referred to
were only feebly voiced and often fricative.
526
J. R. FIRTH —
r or I sound previously mentioned, wHch lie described as a sort of z).
The first is ‘‘ ut r ; dulciter tamen. et tenerrime profertur prope
dentes, iisdem quasi compressis.’’ This is a good enough description
of what we now recognize as a palatabzed r. The second r is " ut
r Latinorum, asperum valde ct durum He also notes the interesting
fact that “ si vero haec littera duplicetur, efformatur duplex ft.
quod effertuT compressis Cjuasi dentibus, lingua ipsos impellente.’’
This pronunciation is also clearly what we now recognize as
palatalized, and, moreover, it is obviously quite a different sort of t
from the dental and retroflex I's which he also describes. It is the
palatalized alveolar t, giving three different places of articulation for
t sounds in intervocalic position in Malayalam : dental, alveolar,
and retroflex.
Of the retroflex t he says : “ est autem Europeis admodum difficilis,
ac pronunciatur inversa omnino retrorsum lingua, adeo ut interiorem
palati summitatem attingat ” — doubled intervocalically “ cum maiori
tamen imjpetu
The dental t and similar sounds were, of course, easily recognized
as Latin. Bearing in mind Peanius’ observations on vis, conalus, and
impetus, and aspiratio, it is interesting to find he regards all the
aspirated voiceless consonants as tense and transcribes them as
double consonants aspirated. Aspirated dental t is romanized as tth
‘■quasi duplex tt cum major! impetu ”. Similarly aspirated p is
“ duplex p cum impetu et aspiratione."
The author's notes on the aspirated voiced stop bh and the dental
dh show he was not merely a slave of the spelling and that he really had
a pretty good idea of the sort of bodily actions that produced these
strange sounds. Of bh he says: “ ut duplex b ] efformatur ex intimis,’^
atque cum vi et aspiratione profertur. " Of ^h : “ profertur ex intimis
cum conatu et aspiratione.” So very few Europeans succeed in under-
standing the single stroke effort ex intimis, required for an Indian bh,
that one feels this insight must have been based on personal knowledge
ex intimis.
Of the dental dh, however, he saj-s : “ quasi duplex dd, cum aliquo
leni impetu,'’ and does not mention aspiration. But for ddh occurring
medially he says : “ idem cum maiori vi, et aspiratione.”
He lists most of the nasals including an a ” ut n Latinorum dare " ;
' “ Ex intimis ” is a very good guess at the motor background of the aspirated
consonants, which are single stroke efforts, the release of the stop synchronizing wdth
a kick ’ of the diaphragm, etc. Cf. ex pectore ” above.
ALPHABETS AA'D PHONOLOGY IN INDIA AND BURMA 527
and another “ ut h, cum aliquo tamen narium ministerio " ; that was
the best he could do about the retroflex n.
He describes the palatal nasal as being like the gn of Italian, but
not identical with it ; efformatur prope dentes cum aliquo narium
ministerio.” This is also in accordance with modern observations.
Other consonants noted are ; —
va, ut u consonans Latinorum, aliquando ut b.
Scia, ‘‘ ut c gallico ore prolatum.” and different from this,
Sza “ inter s et c pronuntiatur, inflexa ad palatum lingua."
Sa, “ ut s Latinorum formatur prope dentes, quasi sibilando."
Kcia, “ retracta lingua et ad palatum inv'ersa cum impetu. et
aspiratione profertur.''
ha ut Germanico ore prolatum.
la — single and double — ■“ est quoddam genus I, quod inflexa
omnino ad palatum lingua crassiori sono efformatur.”
za “ quasi 2 Latinorum, dentibus labiisque vix apertis
pronuntiatur, retracta tantillum lingua.” He clearly distinguishes
s, js, and g — three sibilants.
After all this excellent phonetic description, he concludes :
“ genuiuus enim ipsarum sonus non scriptis, sed voce est aquirendus.”
In the transcriptions at the end he makes use of grave and acute
signs as some sort of indication of accent. Any detailed phonological
study of Malayalam would have to pay special attention to accent
and intonation.
The Alphabetuni Barmanorum seu Regni Avensis was first published
in 1776, but it was much improved in the revised edition published in
1787. It represents the joint labours of Carpani and Mantegazza of
the Catholic JL’ssion. ^
^ iNly attention was first drawn to the Alpkahetum Barmaiionim hy my friend and
colleague, Mr. G. K. Harvey, Lecturer in the Indian Institute, Oxford, who also
very kindly wrote the note on the Mission, quoted below. Carpani knew both Ava
and Pegu, spending seven years in Rangoon. Bishop Percoto sent him to Rome
with ‘‘ accurate information ” about the mission, Burma, and the language. There
is a short note on the Alpkahetum by E. Luce in the Journal of the Banna Research
Society, August, 1914, p. 144.
“The Catholic mission was small but already old when the first Protestant
missionary landed in 1813. Indeed, there had always been a couple of Goanese priests
m Burma from the sixteenth centuiy- onwards, under the Portuguese hierarchy in
India, but they confined themselves to the feringhi colony and were, in addition, only
semi-literate. The first mission, that of the Missions Etrangeres de Paris (now the
dominant Catholic mission in Burma), lasted only four years, 1689-1693, and ended
m martATdom, but it was followed by an unbroken succession of Italian Bamabites,
528
J. E. FIETH —
The AJphabetum Barmamrnm notices most of the outstanding
features of the phonetics of Burmese in presenting the syllabaries
of the Burmese writing lesson. To begin with it points out the
special role of aspiration, glottalization, and nasalization. “ Flures
Barmana lingua habet aspirationes, nasales, gutturales, ahasque.
quibus ea locutio nobis perdifficilis est.”
Carpani not only notices the aspiration of plosives but also of the
four nasals and of I and u'. “ Quatuor priores nasali afficit aspiratione :
qua nempe aer in pronuncianda littera per nares exploditur.' Of
hi and hw he says : “ quasi pronuncietur fla, fua." Moreover,
he draws attention to the morphology in this connection “ In hac
denique lingua per solam .saepe aspirationem significatio activa
tribuitur verbo neutro aut passivo. Sic, kid (ca) cadere, vel decidere ;
kind (cha) deponere vel deiicere ; {hint] dimittere ; \l.ut] liberum esse.
In addition to noticing the antithesis of aspirated and unaspirated
consonants, he fully appreciated similar qualities in the vowels or
syllables. A certain sign, for example, postspirandae syllabae
adhibetur,” which we now call breathy voice and correlate with length
1721-1832. and it is to these that we owe our first studies of the language. There
can be little doubt that both Judson, the founder of the American Baptist Mission
in 1813, who wrote the first great dictionary, and the American Baptists whovse studies
thereafter held the field, were indebted, if only indirectly, to early Catholic MSS.
which no longer survive, the bulk perishing in the fire of 1840 which burned down the
headquarters mission station at Chanthayua in Shwebo district. Within four years of
their arrival in 1721 the Barnabite Fathers had compiled a .small dictionary, and in
the next few decades they wrote MS. grammars and bilingual devotional works,
but the first printed work was the AJphabetum. Its author, Melchior Carpani, who
arrived in 1767 and does not seem to have returned after leaving for Rome in 1774,
was stabbed by one of the Goane.se priest.s, who persistently resented the intrusion
of the Bamabites, men of a high type, whoso mere presence inevitably invited com-
pari.son.s ; his first edition, 1776, was doubtless based on the work of his colleagues,
and the second, 1787, was revised by Mantegazza. Fr. Caejetan Mantegazza, arriving
in 1772. died as bishop in 1794 at Aniarapura, the then capital where his tombstone
still exist."^ : Mhen sailing for Rome in 1784 he took with him two Burmese converts,
one of whom, an ex-Buddhist monk and hence a scholar, assisted in the printing, at
Rome, not only of the Alphabetum but also of a Burmese prayer book, catechism, and
dialogues. Fr. Johannes Maria Rereoto, who, mourned by the author of the Alphabetum
as a better scholar than himself, arrived in 1761 and died as bishop in 1776 at Ava
the then capital — the Burme.se jieriodieally changed their capitals — left translations
of epistles and gospels, Genesis, Daniel, Tobias, i^t. Matthew, prayers, catechism,
etc., and a Burmese— Latin-Portuguese dictionary*, some of which seem to survive in
the Library* of the College of the Propaganda at Rome. J8ee Bishop Bigandet, Outline
of the History of the Catholic. Missiony 1720-1887 j Rangoon, Hanthaw addy* Press, 1887 ;
Hosten and Luce. Bibliotheca Cathohea Birmanica^ Rangoon, British Burma Press,
1916 ; G. E. Har\'ev, History of Burma (Longmans, 1925), pp. 214, 230, 253, 278,
34.5, 349”.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 529
and falling tone. Another is “ signum produeendae syllabae While
of the opposite kind are the signs which mark short checked syllables,
e.g. “ syllabam corripit." “ Punctum suppositum syllahani hrevissinii
reddit soni, et quasi truncat.” In describing vowels he notices an t
which is long and an ?, " breve ac quasi truncuin," and also the
opposite kind of syllable which he transcribes Jixh, “ cum e aperta et
postspirata.’' ^
His account of the pronunciation of syllables written with final
p. t, k, and the check mark is quite in accordance with modern
observations, so that the final glottal stop in such short syllables was
usual in the late eighteenth century, e.g. on a syllable which he
transcribes Jcspk he remarks ; “ ita tamen, ut posterius k vix audiatur :
nempe \fix cnunciari coeptum supprimitur ; quod quidem in qualibet
muta finali observandum est."’ ^
He describes nasalized vowels by comparing Burmese syllables
with French words. Burmese syllables transcribed with final it as
ksen, and kdun he likens to French rin, pain, bon, baton. But he realized
they were really different from these, for he adds a remark which
modern observation confirms : “ w vix coepta supprimitur.'’ On the
syllable transcribed kbuii his remark is also in accordance with modern
observations: “ fi, nasalis. diphthongus vix percipitur.’’ He also
noticed that such nazahzations when followed immediately by the
initial consonant of the next syllable, usually formed a homorganic
junction, heard as m + b, n + d, q + g, etc.
Carpani noticed the behaxdour of the stops in various contexts :
“ saepius vero t, p, aliasque fortes, duplicesve in d, b, et in alias
simplices, seu tenues. Sed quasdam hac in re licuit regulas animad-
vertere, quas in alphabeto notavimus.’’ And later he observes that
after certain nasals and other syllables “ consonantes immediate
sequentes, quae valide, seu durae sunt ut p, I. k, pronunciantur
b, d, g, paucae admodum exceptiones hac in re obtinent."
In phonological terms we should now say that the presence or
absence of aspiration or " breathiness ", is used far more, has more
linguistic weight, than the presence or absence of voice, or the voice
correlation. Whereas the correlation of aspiration differentiates most
^ See my “ Notes on the Transcription of Burmese ", Bulletin S.O.S.^ Vol. VII,
Part I, 1933, also the remarks thereon of Professor Trubetzkoy in his recent Arihiiung
zu phonologischen Beschreibnngen, 1935, p. 29,
2 In Modern Korean the final voiceless stops j?, t, ky are held, and quietly released.
There is no plosion. But they do not seem to have given place to the glottal stop.
530
J. R. FIRTH — •
articulation types in pairs, and this in most of the t}’pical contexts,
there is one very common context in which the voice correlation does
not function, and a second context in which it is doubtful.
The first is the context immediately after very short syllables
ending with a sharp glottal check, where the only unaspirated plosives
to occur in famihar speech are of the p, t, k typie. Taking the bilabial
class of stops, p, py, pw, and ph, phy, phw are all possible, giving
six alternants of the bilabial stop class in this context — and no
further differentiation by voice. Here we have what I have termed
uncorrelated p, t, k.
Something very like the opposite would appear to be the case in
the context immediately after long syllables with closing nasalization,
where in most cases, but apparently not in all, the sounds heard are
like b, d, g, and are not used in contradistinction from p, t, k. These
I should term uncorrelated b, d, g.
It seems to me quite unnecessary and probably erroneous to
postulate relations between the stops in these two utterly different
contexts. The question of notation or what letters we shall use in
Eomanic orthography is another matter altogether.
Another feature of the Alphabetum which deserves notice is the
classification of the different t}pes of syllable to be met with in this
so-called monosyllabic language. It seems to me more enlightening
than a mere catalogue of so-called indi\idual sounds, perhaps because
it follows the Burmese traditional writing lesson, and also because
it agrees in some measure with the modern contextual approach.
He gives six classes of syllable in the orthography, but naturally
some of these classes correlate also with phonetic habit and
morphological structure. There is, of course, the distinction between
the short sharply checked syllables and the long breathy ones, which
we have already noticed as two characteristic contexts in which
immediately following consonants should also be studied. He also
notices those which begin with aspirated consonants, as well as
the breathy ones which fade out, “ postspirandae.” Then there
are the diphthong-syllables. “ Diphthong! autem, atque etiam
triphthong! in tota lingua Barmana frequentes admodom sunt.”
He notices ei with “e praestricta ”, ou with “o medium ”, and
also ai and au.
There is another characteristic contrast of syllables in Burmese
between those which begin with a yotized consonant group and those
beginning with a labio-velarized group : e.g. py, phy, my, hmy, ly.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 531
hly, etc., against pw, phw, mw, hmw, Iw, hlw, etc. These form
classes five and four in the Alphahetum. Of the yotization of
syllables, it says : “ Hoc igitur signuni brevissimi sonuni i syllabae
intrudit, ut est in ghiaccio, pianta apud Italos." The w ” sound in
the other class is given as the u of the Italians or the ou of the
French.
The sixth class is really only due to orthographic superfluities, the
Burmese letter “ r for instance, being pronounced “ y
Carpani’s description of the vowels holds good to-day. Barmani
septem habent sonos, seu vocales Italorum : duas e ; apertam nempe,
et praestrictam : duo o ; medium, et largum, seu apertum : atque
a, i, u.’" Failing ordinary letters for the two extra vowels he employs
se and the Greek to in his transcriptions of the open e and o.
The close e “ ut in nee, portee apud gallos.
The open se ut e in apres, or chaine, grile apud Gallos ".
For the open o he suggests “ apertum ut in veto seu vuoto apud
Italos ”.
For the two letters corresponding to y and r he gives the same
pronunciation. His note on r being " quam multi in pronunciatione
in precedentem mutant
For the Burmese characters which may be transliterated hy,
hSy, and hr he gives the Italian indication scia or French chien. This,
too, accords with modern observation. In Romanic orthography this
element could be written hy, as it was probably an aspirated yotized
group originally, and is now pronounced J, rather like a certain very
fronted pronunciation of the fcA-laut.
Carpani shows acquaintance with French, but not with Spanish
or English. He found the velar nasal q difficult to describe.
He says of this letter ; “ quam per ng utcumque expressimus.
simplicis est soni, nullisque nostris litteris e.xprimi potest.'’ And
the best he can do for the sound 6 is to suggest it is a lisped s.
" absque sibilo ; uti apud nos quoque in nonnullis auditur \dtio
linguse vel educationis.”
He heard a, g, '' iu.xta Germanorum pronunciationem," and the
affricates c and j he transcribes as ts and tzh. On the whole, it will
be agreed this Alphahetum was an excellent piece of work for that
time, and was not surpassed or equalled until our own day.
The following table of letters is appended as an example of a
schematic Romanic alphabet for Burmese : — •
532
J. E. FIETH —
Example of Oxe Aeticulatiox Type
Initial Alternance Onlv
Bilabial Articulation involving :
lip closure Alveolar Velar
Basic type of
articulation
Voiceless
' [P]
Aspiration
Voice
Correlation
t
1
k
P
ph
b
1
t, etc.
k
Yotization .
py
phy
by :
nil
nil
Labio-velarization
pw
phw
bw
tw, etc. !
kw, etc.
Nasalization
hm
i '
1 m 1
hn, etc. '
hi), etc.
Nasalization and
Yotization
hmy
my 1
nil
ny or special-
ized palatal p
Nasalization and Labio-
velarization
hmw
mw
hnw, etc.
r)W
Total
9
6
10
9
Total of 15 “ substitution
counters ” or terms having
basic lip closure or bilabial
articulation occurring in initial
position.
Total of 34
for plosives
and nasals in
initial posi-
tion.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 533
Schematic Alphabet for Burmese in World Orthography
Initial Alternance Only
Bilabial
Articulation
Alveolar
Articulation
1
1
Pre-palatal Palatal Velar Glottal
Articulation Art. , Articulation Art.
p ph b
t th d
c ch j 1 k kb g
Yotiza- , ,
tion Py
: (see c, ch, j)
Labio-
velariz- pw phw bw
ation .
tw thw dw
cw chw jw kw khw gw
; Nasals hm m
hn • n
hq g
i Yotized ,
’ nasals ^
iiilSdlS
f
(see hny, ny)
Palatal
hny ny ■
Labio-
velarized |
nasals hmw mw
w
hnw nw
'
gw
Dentals |
(6 8 -
i 0w 8w t
bi 1
hly ly
hlw Iw
1
)
-Laterals similarly treated.
)
s sh z J
shw ‘ zw I
A
w
'
y
h
1
i
yw
>4
(hy = J)
534
J. R. FIRTH —
Yowels.
Low level tone, long, greatest
i
e
E
a
0 0
u
frequency of occurrence.
eig
ag
1 oug
ug
aig
aug
Falling tone, long, breathy
'i
■ 'e
'e
1 'a
0 0
'u
'ig
'eig
ag
'oug
'im
'aig
'aug
Slightly falling, medium length.
i'
e'
e'
a'
o' o'
u'
creaky voice, weak closure.
ig'
eig'
ag'
j oug'
ug'
aig'
aug'
Very short, “ bright voice.
i?
, ®i^
E?
a?
ou?
u?
ai^
au?
abrupt closure, slightly falling.
Neutral .... a
i '
TEXT IX BROAD TRAXSCRIPTIOX
I. Burmese
0a 'dii] za sha ya hniq* 'nyii] 'khouq 'jiq
lu ta yau^ 9i, 0a 'diq za da zau^ go, kaig 'pyi hlyiq, 6a 'dig za
tai^ 0o' '0wa yau^ ywe', 0a 'dig za sha ya 'ji 'a, ci' 'zag. Big do'
0a 'dig za ha, 'kaug 'la ; cou^ a myi go lu 8e sa 'yig 'de ma, ths'
'tha dE. ci' 'zag sho 'pyi, 0a 'dig za go pya' ywe', 'kaig, da be'
ue' 'pya jig '8e Sa '1e, hu', 'do 8a' hnig' 'me Si.
6a 'dig za sha ya 'ji ga', o, '0i khag da mu ba; di ha, sa 'loug si
Sa 'ma ga 'le 'mya, 'hma lo', phyi^ pa leig' me. '0i khag do mu ba,
hu', yo Se zwa 'taug bag 'to '/o 'pya Si.
lu 'ji ga', tE 'kaug de. Big do' do', sa 'loug 'hma 'Swa youg. cou?
hma bE poug a 'co you^ mE sho da, ma 'pya naig 'bu, sho lyE?, 0a 'dig
za sha ya 'ji ga' kha' ga' ya' ya' 'taug bag ywe', ce ya' So 'li, 'do 8a‘
ga' ma pye Si' lE^kha na hnig' 'pyo Si.
0a 'dig za sha ya 'ji ga', a 'co ma you? ya' aug, cug do pyig ba'
mE. UE? phyag Sa 'dig za hnai?, 'mwe 'phwa sa 'yig ga', khig 'bya a
myi go, tha' lai? pa' mE pyag 'pyo le 8i.^
' Story from Armstrong and Pe Maung Tin’s Burmese Header, p. 41. Recorded on
H.M.V. Cl 181.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA
535
TEXTS IN WORLD ORTHOGRAPHY
II. Tamil
paal ku|;ikkaata puunai un^aa entru raajaa tennaali raamanijam
kee|;|;aar. tennaali raaman ui|_Ju entru patil connan. appa^iyaanaal
koit^uvaa entru uttaravi^aar.
tennaali raaman tan viif^uku pooy puunai ontrai pi|;ittu atan
munnaal kotittukkoi|^|;irukkum paalai vaittaan. paalai ku|^ikka
poona puunaiyin mukam ventu pooyittru. ippu|;i iraii_|;orutaram
na|;anta pin, anta puunai eppootu paalai kan|;aalum (ojo^Jam
pi|;ittatu. ka|;aiciyaaka puunaiyai raajaavi|;am kaaij, pittaan. raajaa
mikavum aaccariyappa]]|]aar.
III. Telugu
voka roozuna tenaaliraamalingani kot^uku tondara tondaragaa
raazugaari darjanaaniki vaccinaa^u. kai]_(^|_anta niirugaa vaccinaai^u.
vaanini cuucina vaarandariki aajcaryamu pu|;|;inadi. ataniki eemi
aapada vaccinadi janulaku booda pa^aleedu. vaai^u tinnagaa
raazugaari vaddaku vacci mahaaprabuu !
maa kompa munigipooyindi. naa tandri canipooyinaa^u ; ani
manavi ceesukunnaa^u. aa sangati vini raazugaaru aajcaryapa-
<^inaaru. aayananu aajrayincukoni, tannu dukhamulandus antoogape^u
cundina raamalinga^u gatincanandoku raazugaaru mikkiU yyasana-
pa^aaru. veii|;anee aayana, tana paricaarulanu pilici, konta
dravyamunu tenaali raamalingani komaarunikiwamani uttaravicci
naaru.
IV. Marathi
eka baiyne ek mwggws paj^yle bote, tya mwqgsala tytsa far la^a
laglela beta, te tycya gbari ekadya mwlapramaiie vavrat ase. byamwl_e
tya gbarat sapakyr^aci agdi bbiti nase. eke dyvji ase dzhale ki, ti
bai aplea tanbya mwlala badzevar nydzvun pai|,i ai|,avayas nadivar
geli ; ytkyat ek mo|;ba sap gbarat Jyrun, tya badzevar tsacj^bu lagla.
mwqsaci drg|;i tyadzvar dzatats, mwlala ba tsavqar ase pabun tyane
tab^tob sapavar w^i gbatli, an,i tyaci kban,^o[i karun takyli ! mwlaci
ai gbari yete to daratats mwggws tyca drg|;is pa^le; tyatse toii_(^
raktane bbarlele bote, apla parakram tyla dakbvyij^ya karytats
dzan_w kay te tycya ton^aka^e pahat rabyle ! paij, tyla ase va^le ki,
mwqsane az aple mul marun kballe ! tevba tyla dzo rag ala tyacya
536
J. B. FIRTH—
dzli3pa|;yat tyne bharlela hai^^a mwqsacya dokyavar |;akyla. tya
mu[e mwqsatse ^oke fu|;un te tatka|_ maran pavle. pw(^he ti bai
badzevar dzaun pahate, to sapatse twk(^e tycya drgjis pa^le, va mul
hi khw/al nidzlyatse tyla a^hal^un ale.
V. Urdu
ek ba|;he admi ki mawtka vaqt jab qarib aya to ek lak^idka
ga];tha bandkar apne bej;6 se kaha ky wse toi;e. har ekne bari bari zor
lagaya lekyn gatJha na tu|;a. bufhe tajrubakar bapne phyr ga^heko
kholdia awr ek ek karke sari lak];ia tordi. phyr apne be|;6se mwxatyb
hokar kayhne laga “ pyare be^o ! jab may ys jahase kuc karjau to
twmhe cahie ke ys lalq;i6ke ga|;t;he ki tarhe raho. agar twm baham
pyaromwhabbatse rahoge to taqatvar se taqatvar admi bhi twmhe iza
na pahwca sakege. lekyn agar twm ek dusre ke dwjman bangae awr
alayhyda alayhyda hogae, to kamzor admi bhi twmhare ys nyfaq se
faeda w(;hakar twmhe zarur nica dykhaege.
VI. Modern Persian
yeki ®z ae^yane maemlekmte iran, do mdsed mnjir ke taze dmr
bay®/ rmside bud, bsraye pade/ahe xod be^envane tohfe f®restad ve
d®r mrizei ke be/ah nevejt motezmker Jod do ®d®4 asnjir ke pijsz
mousem rmside bud, tsqdime m^lahmzrste homayuni mikonaed. Jah
paesmz xandmne morasele xeili xoj’vmqt gmrdide. vmhokm kmrd ke
qasede hamele srize mnjirhara behozur avmrde en^ame xodra
begirffid. vmli mote^mjjebane did fmymt yek mnjir dmr z®rf®st mz
qased porsid an mnjire digmr kn, qased ®rz kmrd an digmrira
xord®m. Jah moteymyyerane goft cetour xordi, qased ®rz kmrd in
tour, ve qoulra ba fe^l yeki nmmude mnjire doyyomira bedmhmne
xod gozajte v® xord.
After the presentation of a schematic alphabet for Burmese with
an illustrative text, and experiments in world orthography for Tamil,
Telugu, Marathi, Urdu, and Persian, attention must be drawn to
certain fundamental questions raised by the employment of world
orthography to s}Tnbolize the forms of languages which have hitherto
made no systematic use of Romanic characters.
The first thing to be said is that the Roman alphabet has been
found to work w'eU from the days of greater Rome to the present time.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA
537
wlien Western civilization is become a world civilization. Hence the
phrase world orthography. The Roman alphabet has proved practical
in all kinds of printing, both by hand and machines. In education in
its widest sense, in all manner of notation, popular and scientific, it
serves us well. We have evolved a variety of founts of tvpe and
spacing for effective lay-out in all sorts of printed language. We have
developed new letters in harmony with the alphabet, also accents and
punctuative signs. No nation, no people, need he.sitatc to adopt it.
Those who have will not go back.
On the general advantages of the Roman alphabet, Professor Otto
Jespersen has written a useful article which serves as the introduction
to a report published in 1934 by the League of Nations Co-operalioii
IntellcclueUf. entitled '' L'adoption universelle des caracteres latins’'.^
Of the practical advantages, one or two forceful illustrations mav
be given. The technical re.sultsof Romanization in Turkey are : With
Arabic characters a compositor could handle 4,500 in six hours. With
Roman 7,000 in the same time. The cost of production has been
reduced from 25 to 50 per cent according to the size of the work. The
number of touches on the typewriter keyboard has been reduced from
90 to 37, and the employment of machines and typists enormously
increased.^
Professor van Ronkel, of the University of Leyden, writes ^ :
“ L’ecriture javanaise est belle, mais compliquee et peu economique :
un text y occupe trois fois plus de place qu'en ecriture latine. Les
livres sont done trois fois plus gros qu'il n'est necessaire.”
Furthermore, the Turks have proved the great advantages of the
new alphabet in schools, and a rapid multiplication of books and
libraries has followed the adoption of an alphabet at once simpler
and better suited to represent the forms of the language, and also
much cheaper in production co.ffs.
In addition to practical advantages of this kind, the Roman
alphabet has definite merits as the framework of a scientific linguistic
notation. It lends itself to anali'sis and svnthesis. It does not build
svllabaries. It is anaUffic. using a comparatively small number of
signs which can be arranged and employed to suit the phonologc' and
morphologv of almost any language. Moreover, the synthesis of the
^ Afterwariis reforrcil to as A.U.C.L.
- .See A.U.C.L., pp. 126-0. by Professor Caferoglii. of Constantinople. And p. 136,
by Profe.ssor Rossi, of Home.
“ A.U.C.L., p. 92.
VOL VIII. p.\RTS 2 .cxn 3.
3.1
538
J. R. FIRTH —
letters produces easily recognizable ditferentiated word-forms as wholes,
the differential elements suitably symbolized by letters or signs having
their places in the word and also in an ordered series of alternants
established liy analysis. Having analysed the language into a number
of ordered series of letter-units, you put the " pieces '' together again
and find you have differentiated word-forms. When vou put together
your pieces '' and find the result corresponds with the facts you have
a scientific or “ organic ” alphabet. It is not surprising, therefore, to
find Professor Caferoglu reporting that the adoption of Roman
characters “ a conduit egalement a une simplification de la langue “.
that it has had some influence “ sur la grammaire turque and that
it has opened "de nouveaux horizons aux recherches philologiques
concernant la langue turque.” ^
The alphabetic revolution in Turkey is. in fact, the most significant
movement in the recent history of the world alphabet. It was at the
epoch-making Turcological Congress, held at Baku in 1926, that
representatives of the Turki.sh republic, of the Turco-Tartar peoples
of Russia and of Ru.ssian and other European Universities, proclaimed
the necessity and the opportunity of abandoning the Arabic alphabet
in favour of the Roman character. Within two years the alphabetic
revolution affected about 25,000,000 Turco-Tartars in the U.S.S.R.,
which has been carridng out a great Romanization programme ever
since among the Asiatic daughter republics.^ Turkey herself followed,
and what is called the N.A.T. (” nouvel alphabet turc ”) came into
force partially on 1st January, 1929, and completely on 1st June, 1930.
In Rus.sia it is true that earlier moves towards Romanization had
been made by several local governments, in particular by Azerbaijan,
but the Congress of 1926 marked the beginning of comprehensive
alphabetization " on a vast scale. In 1922 only two languages of the
Ru.ssian Union had adopted a new alphabet, but the number given
for 1933 is seventy !
The Russians attach great importance to the unification of the
many Roman orthographies which had been independently devised
in pre-Soviet days, and they report : “ Actuellement I’unification se
poursuit ; elle se manifeste par la progression de la fusion phonetique
et graphique dcs alphabets nationaux, la simplification de la fortne
des lettres et la reduction de leur nombre.”
' Ste A.U.C.L., pp. 124 5. Italics Professor Caferogluts.
= See A.U.C.L., pp. 1.13-4, by Professor Ro.s.si, the official Soviet report on p. 161.
and a report on Romanization in the U.S.S.R. by Profes.sor Braun, of Leipzig,
pp. 142 seq.
ALPHABETS AND PHONOLOGA' IN INDIA AND BUKMA 539
That leads us to the elements of the whole technique of symbolizing
the forms of a language by the use of Roman letters. It may be
described as alphabetical economy. To make the most economical
use of letters it is essential that the fulle.st advuintage be taken of
contextual conventions, thus reducing the number of signs required
not only for the symbolization of the terms of an ordered series of
possible alternants in any given context but for the particular
orthography' as a whole. To achieve this, thorough phonological and
morphological analysis is nece.ssary'. '' Si toutes ces questions ne sont
pas suffisament examinees et resolues d une maniere uniforme.
I'orthographe reste maladroite, lourde. difficile a lire, et elle a peu
d'attrait pour les indigenes, qui doivent d’abord s'habituer a la lecture.
Sous ce point de vue les orthographes deja existantes ne sont certaine-
ment pas toutes tres satisfoisantes.’' '
To a nucleus of thirty-three Roman letters the Soviet linguists
have added fifty-eight new ones, making a total of ninety'-one letters
in a sort of unified alphabet which they think will serve all the languages
of the Union. There is a great danger of swamping the characteristics
of the alphabet if too many new letters are employed. This may' quite
well result if the letters are ba.sed on universal phonetic categories
instead of on a phonological analysis of each language ad hoc. The
Arabic and Indian alphabets are such that they' have developed either
initial, medial and final forms or special compound letters. Such
specializations of form may' even be ju.stified by abstract general
phonetic theory, but very' little can be said for them from the
point of view of alphabetical economy. An orthography can be too
phonetic. The value of a Roman letter depends on its position and the
context.
A certain number of new letters such as those devised by the
International Phonetic A.ssociation are undoubtedly' necessary'. But,
as Professor Troubetzkoy' quite rightly' points out, " Souvent ces
caracteres modifies sont tres nombreux de sorte que I'aspect general
d'un texte ecrit dans un tel alphabet est tout a fait ‘ exotique ’. En
raison de cette circonstance un des principaux arguments cite
d'ordinaire en faveur de I'adoption des caracteres latins pour toutes
les langues du monde se trouve presque reduit a neant." - Most
spelling is phonologically. not phonetically', representative.
A schematic sy’stem of spelling or regular alphabet which enables
^ A.U.C.L., p. 34. Professor D. Westermann reporting on Africa.
- A.U.C.L., p. 48, reporting on the peoples of the Caucasus.
540
J. R. FIRTH
US to sjTubolize the forms of a language by means of combinations of
letters and other signs Avithoiit redundancv' and vet without ambionitv
must be based on linguistic analysis and involves the consideration of
word formation and sentence structure as well as of pronunciation.
From the Saussurean point of view, which has been applied and
developed in English by Dr. Alan Gardiner/ orthographv is
representative of language, not speech.
That is a striking way of saying half the truth, and perhaps the
half that has too often been obscured. But there is quite obviouslv
a danger in following Baudouin de Courtenay, de Saussure and
Durkheim to the extent of the abstract integration of '' sounds ’’ or
■■ phonemes ’’ or letters and signs in a mental scheme of ideas or in
"the language as a whole". This kind of abstraction goes further
than is at present either nece.ssary or desirable for the handling of our
facts. In the sj'mbolization of the forms of a language by means of
an ordered system of letters and signs, the first principle should be
the recognition of characteristic recurrent contexts in which an
ordered series of phonological substitutions may take place.
If we take an ordered series of English words or forms such as
bi'.rj, bid, bed, beed, ba'.d. bo:d, bind. bid. bi'.d beid. bond, bn id, baud,
bold, bud. beid. we have sixteen vowel alternants in what may be
considered the same context. Between d and g, however, only three
are possible.
In final position tiie number of possible vowel alternances is two,
.sometimes three less than in medial position. In initial position in
i.solated words the nasal alternances arc m, n, in intervocalic and final
positions m, n, q. But immediately before a final k onlv q is possible,
before final p only m is possible, though before final t and d, which have
morphological function, all three are again po.ssible, m, n, q.
In the application of World Orthography to Indian languages the
letters m, n, p, q, have been used to represent the unique homorganic
nasals preceding certain stops, as well as for the .series of nasals which
may occur in initial position. But we refrain from any functional
identification for example of a .s])ecific or unicpie tn on-glide to a
homorganic .stop, and an m as a term in a three, four, or five term
alternance in initial position.
Tlie initial medial, intervocalic and final positions in Tamil
agglutinations give contexts in each of which various series of
^ In hi.s Theory of Speech and Language, Oxford. 1932.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA
541
alternances may take place. These must be studied in close connection
with the morphology, each series of terms in each context
independently, at any rate in the first instance. Similar considerations
apply to Malayalam. In that language the consonantal alternance
in initial position includes, for example, two homorganic plosives
differentiated by the voice correlation which we may symbolize by
k and g. In intervocalic position, however, the consonantal alternance
includes two homorganic sounds differentiated by the tensity-laxity
correlation which appears to affect the whole manner of articulation,
involving also length, and a parallel laryngeal correlation of some sort ;
both these are again differentiated from a third sound by the
voice correlation. This gives us a three-term alternance which we
may sjmibolize by means of kk, k, and g. It so happens that
intervocalic kk sounds rather like initial k, and intervocalic k
rather like g, except that it is feebly voiced and often fricative.
It will be seen at once that from the logical and functional
point of view it is impossible to identify the terms of the first
series with the second series. It is practically convenient to use the
same letters over again both from the point of \-iew of pronunciation
and alphabetic economy. We may write toikt and sips, using t and s
both initially and finally, and they may correspond to similar sounds,
but the two f's and the two s's are phonologically and morphologically
different. Or take the English word stick, which may be transcribed
stik or sdik, according to the nature of the contextual conventions
laid down. Discussions have taken place on the further and quite
gratuitous question of whether ” the sound " after the s is to be
identified with t’s or d's in other contexts.
In other words the value of any letter is determined by its place
in the context and by its place in the alternance functioning in that
tvpe of context. This I have called its minor function, but grammatical
and semantic function must also be considered. These I have termed
major functions. ^
These ordered series of alternants vary from context to context,
so that minor function is not a constant for the language as a whole.
The number and nature of the terms of such series also vary from
context to context, ainl it is useful to note the range from contexts
of maximum alternance to those of minimum alternance. i It follows,
’ niv “ U.-'D and Distrdnition of Certain Enj;ii&h Sounds”, English SttaUes,
Kebniarv. IbSo. Also my " Technique of Semantics in T nf thr Philohnx nl
Sorif-fy oj diffit Britain, 1933.
542
J. E. FIRTH —
therefore, that the differential function of the signs or letters varies
from context to context. The same letter may be used to symbolize
terms in several different ordered series of alternances. Its phonetic
value categorized by perception or physiological phonetics may or
may not be similar. A balance must be struck between the con-
venience of using the same letter for terms which are roughly similar
phonetically and the great alphabetic economies rendered possible
if the same letter can be used for a variety of purposes according to
context.
The most uneconomical, I almost said extravagant, alphabets
are those of an abstract schematic order, universal, purely logical
and symmetrical but extra-linguistic. Such alphabets are some-
times necessary for dialect and comparative work and in the earher
stages of phonetic research. But they are quite unsuitable for
descriptive grammar or as a basis of a practical orthography.
The great advantage of this alphabetic economy based on the fullest
use of contextual conventions is what may be described as free letters.
Such redundant letters not required in any particular context may be
used in all manner of ways. For example, in the Burmese orthography
here suggested it would be possible to eliminate the mark which
distinguishes the long falling tone with breathy voice, now written
for example 'la, and use a final letter h, thus lah, but this would
have the inconvenience that many syllables would have to be separated
either by spaces or joined by means of hyphens. Otherwise in
compounds printed together there would be confusion with such terms
as hi, hm, etc.
Other common redundancies arise in contexts where certain
differentiations do not occur, such as the absence of voice correlation
or of distinctions between / and h, r and I, w and v. Sometimes a
redundant letter may be used with the purely lexical function of
separating homophones. For example, in Cambodian, w'hich employ's
an alphabet of Indian origin, the final aspirated consonant is really'
redundant from the phonological point of view. But it serves a useful
purpose in separating, for instance, “ duh conserver, de dulch
malheur Something of the same sort would probably be necessary'
in a reformed spelling of English.
It is probably true that there are no qualities in any letter taken
by' itself w'hich make it inherently superior to any' other. "WTiat matters
^ A.U.C'.L., p. 43, by M. Martini, nf Pans.
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA 543
is again the clearness and distinctness of the differential features.
g is probably better than g, if q is used in similar contexts.
Another important question affecting alphabetic economy and
phonological theory is that of consonant groups and consonant
junctions, which must be clearly distinguished in all phonological
analysis.
Both digraphs and trigraphs are used in my Burmese orthography,
but these compound letters are to be understood to represent not two
or three substitution elements in the forms of the language, but single
substitution counters or terras belonging to an alternance occurring
in initial position. The bodily actions corresponding to these units
are all, so to speak, single stroke efforts. Many of the theoretical
difficulties of phonetics have been due to the mistaken notion that the
events of a phonetic sequence correspond to the string of Eoman
letters used to symbolize the linguistic forms in the sequence.
Some people are of the opinion that one element should be repre-
sented by one letter, not a group of three. As things are I prefer the
ordinary world alphabet as far as possible. But I am comnnced that
we must not allow the characteristics of the Roman alphabet to dictate
the course of linguistic thought.
Consonant groups, such as st, str, sp, spl, sk, skr, in initial position
in English, are best regarded as group substituents, and no attempt
should be made to identify the function of the letter t ” (here part
of a digraph or trigraph) with that of a similar letter used in another
context. It is important however to distinguish such groups from
consonant junctions, cf. . . . missed riding, . . . Miss tried, . . . in my
stride.
The contextual study of such consonant groups and consonant
junctions is likely to produce interesting results from several points
of \iew, phonetic, morphological, syntactical, and also what I have
termed phonsesthetic.^ There are also ob\'ious historical advantages
in this way of regarding groups like sp, st, sk. I have presented a table
of such initial consonant groups in English in the article referred to.
From the foregoing summary of a technique of contextualization
it ivill be clear that no attempt is made to establish psychological or
phonological relations between terms of different series. The contexts
can be systematically analysed and various alternances constituted,
but it does not follow that all these alternances or systems should be
1 See my “ Use and Distribution of certain English Sounds ”, English Studies,
Feb., 1935.'
544
J. R. FIRTH —
forced into a single theoretical architectonic schemed What letters
are practically convenient in orthography is a different Cjuestion
and involves additional criteria. In the contextual technique
I advocate, the statistical method is the one to be followed, and this
allows discontinuity and change of measure and value from context
to context.
If the ultimate units of linguistic material be treated in this way
context by context, there is no fun left in the notorious question
" Are q and h the same phoneme in English ? ^
Further progress in phonology will depend on the constitution of
alternances the terms of which have differential values in the
characteristic and significant conte.xts of a given language. Up to the
present no such exhaustive study has been completed, so that we are
not really in a position to examine what relations, if any, there may
be between phonetically similar terms of different alternances. We
are. of course, accustomed to refer to the influence one " sound
is said to have on another, to inter-syllabic relations such as vowel
harmony, inter-word relations such as assimilation, and at first sight
it would seem that these facts are overlooked in a narrow contextual
technique. They are approached in a different way, and will be
more fully understood when exhaustively examined context by
context.
The minor function of an alternant, that is of one term of an
alternance. is determined by the constitution of the alternance as
a scries of term.s having differential values in a certain type of context.
Other facts are irrelevant. A term is to be considered first in relation
to its context and secondly to the relevant linked alternance. What
relations it may have to the language as a whole is difficult to guess.
To treat a language as a sort of unity does not mean that every element
is to l)e regarded as in e((ual relation to every other element. The
phonological description of a language will reveal not just one
^ 111 the Alphahft}i}ii Bnrmannrum it is obvious from the way various tyj)e> nf
\vllablc3 are presented in the traditional Burmese way, that the number and nature
of the terms or possible sub.stituents ’’ varies from context to context, and that a set
of letters is not being set uj) as a functioning system in vacuo apart from context.
Nevertheless, Carpani linds it necessary to i>>ue the following warning:
” Observandum tamen e‘'t non omnia qunleiii haee signa cum qualibet littera, aut
syllaba coriiunni vel solere, vel etiam pfisse.” That ho should have gone out of Ins way
to say this .'shows that he realized the loinmoii mistake of regarding a set of letters as
a whole as free units ttr tcriii'^ in a sort of mathematical relationship.
See Twaddell, On defining the Phoneme,” Lnnjjuage Monograph -Vo. -VI'/,
pp. 10 ff. and 25 ff.
ALPHABETS AND PHONOLOGVT IN INDIA AND BURMA 545
architectonic system, but a series of systems which taken tofiethei
give a complete and unambiguous account of the facts.
In the translator's preface to Holger Pedersen's Linguistic Science
in the Nineteenth Centunj, Professor Spargo calls the reader's attention
to one important feature of the book — “ the striking role assigned to
the study of phonetics in increasing our knowledge of linguistics. It is
shown clearly that every important advance during the last century
and a quarter was made by a scholar who attacked the problem from
the phonetic side." During the last twenty years phonetics ha.s been
applied in all sorts of practical ways. One of these has been the
establishment of orthographies for hitherto unwritten languages, and
of simple, readable unambiguous transcriptions of languages having
either an unfamiliar script or one which docs not correlate with the
forms in actual use even from the native point of view. Considera-
tions which help us to establish such a notation are technical and
practical, and cannot fail to have a profound influence on the future
of linguistics.
Further, and perhaps most important of all, those of us whose daily
business it is to study the speech behaviour of our neighbours without
either envy or scorn, and also that of strangers without breach of
courtesy, realize as no one else can how narrowly conditioned our
speech habits are by the daily round, the common task. Within the
framework of social routine and the ritualistic give and take of con-
versation there are great ranges of possibility, but few are unexpected.
If you disturb the air and other people's ears by using your speech
apparatus in ways both unexpected and highly individual, you run
grave social risks.
Such behaviour, to say the least, is felt to be unusual. It is
generally tactless, though .sometimes it is merely eccentric.
Occasionally we condemn such behaviour by sapng it is uncalled for.
or more strongly by saying it is not done. We usually have a cue for
what we sav ; the lines too. are there, and though there may be a choice
our fellow-countrvinen know them and know what to expect.
To stretch the metaphor, what we say is usually "called for".
With the linguistic stranger things are different. If you are wise you
will be prepared fur anything, do what is practically convenient from
moment to moment, and avoid strain or weariness, by reducing the
neccssitv of vocal interchange to a minimum within the bounds of
international courtesy.
For some years now I have stressed what a friend and former
546
ALPHABETS AND PHONOLOGY IN INDIA AND BURMA
student calls the Handlungsckarakter der Sprached and also the very
fine distinctions in speech behaidour, determined by tj-pical recurrent
social situations for which these locutions are specialized and of
which they are organs or functions. It follows from this, of course,
that a great deal is demanded of our notation and descriptive technique.
Without it accurate morphology is impossible, and without
scrupulously identified forms and well estabhshed texts Semantics
is apt to be just gossip.
It is the first duty of a describer of language, as it is of a classical
philologist, to establish his forms and his texts with a scrupulous
exactitude. For what is the semantic value of a corrupt text ?
The purpose of this digression into general hnguistics is to show that
not even the broadest explorations in sociological linguistics are hkely
to lead to sohd results without the pedestrian technique of the ABC
as the principal means of linguistic description.
And of all A B C's the Roman is the best. Perhaps Lenin was
right when he said to the President of the Pan-Sovietic Committee
for National Alphabets : “ La latinisation, voila la grande revolution
de r Orient.” ^
* See a review in A’eae Schwener Rundschau, July, 1935, pp. 17d-8, by Fritz
Guttinger. '■ Zu den nachhaltigsten Eindrucken, welche man von der program-
matischen Schrift J. R. Firth's uber den Sprechvorgang, wie auch von seiner
Lehrtatigkeit am University College London davontragt, gehort die Einsicht, dass
die Spielregeln der Sprache und des iSprechens im Grunde etwas viel Roheres sind,
als man zu glauben gewohnt ist. Was fur Folgen dies fur die allgemeine Sprachtheorie
hat, braueht hier nicht ausgefuhrt zu werden. Daraus, dass das Zweekhafte, Hand-
lungsmassige der Worte und Satze zur Betrachtung abgesondert wird, ergibt sich
letzten Endes die Notwendigkeit, die Formenweit nach streng formalen Gesichts-
punkten zu beschreiben . . .”
2 A.U.C.L.. p. 174.
Mittelpersisch venok „ Erbse (Linse ?) “
Von Bernhard Geiger
T\AS Wort kommt ini Frahang I Pahl., iv, 2. in der auf die
Getreidearten Weizen, Gerste, Hirse folgendeii Reihe der
Hiilsenfriiciite vor ; Seine Lesung. Herkunft und
Bedeutung sind aber bisher nicht festgestellt. Das Pahl.-Pazand
Gloss, von Jamaspji Asa-Haug bietet vinuk mit der unrichtigen
Bedeutungsangabe a seed, grain for agricultural purposes'*. Sie
beruht auf der irrigen Annabme Anquetils (vgl. auch Justi, Bund..
p. 267 3), dass iui Frahang das Wort
falsche Schreibungen fiir = aram. X1‘2 ,, Samenkorn “) noch
(alles
als Erklarung zu dem vorhergehenden gehort. Die traditionelle
Aussprache von ist vlnag, vlnitg, vlnu, linl (Junker, Frahang
115 sub Uui dasselbe Wort handelt es sich natiirlich in der
Stelle Gr. Bund. 117, 2 f., wo in der Klasse der Getreidepflanzen
nach (Weizen), (Gerste) und (nicht Reis, sondern statt
Hirse : Frah. i Phi., iv, 1 ; Bund., ed. Justi 64, 17 vgl.WZKM,
26, 1912, 300) noch aufgeziihlt werden :
Darin entspricht dessen ) hier zum Wort gehbren kann und nicht
■■ und ” bedeuteii muss, dem des Frahang, wie dem
Worte ^^“1 milj daselbst. Das mp.
naxi'dt bezeichnet ebenso wie np. niixud die Kicher-Erbse, und^!^"l-3"'^>
das erste Wort der Reihe, ist mit np. girgir identisch, das nach dem
Burhan eine Feldfrucht. rund, von schwarzer Farbe und kleiner
als nuxud ist. nach anderen aber eine Art von Bohnen (hahld),
arabisiert , nach dem Farh. §u'uri dagegen gleicharab. jLU-,
also eine Lathyrus-Art. ist. Die zwischen und stehenden
Wbrter nicht Xamen von Hulsenfriichten, sondern
548
B. GEIGER
von zwei Hirsearten. Dem Weizeii (-^), dem rat (dein ]\Ieister, d. i.
dem Klassenobersten) dcr grobkdmigpii Gctreideartea, wird Bund.
58. 12 ff. hulnm {ahim. 1 hatnenllc ke j;5 ( J^)-ic -x^ranend ,, die
Sonimeriiirse [halum, ahun, gr. eXvfios, . s _\ t . DI'T'm’
Low. Aram. Pfl.. p. 101. Burhan jjjj die man aucb
j;r (= WZKM.,. l.c.) nennt als rat der kleinkbrnigen Getrei-
dearten gegeniibergestellt. In der cntspreclienden Stelle des Gr.
Bdh. 121, 13 f. steht an Stelle von (,;2). und die.ses ist
offenbar identisch mit^-*Oy in der Verbindung innerhalb
der uns beschtiftigenden Reihe von Hiil.senfriichten (Gr. Bdh. 117, 2 f.).
Diesc.s bezeichnet also cine zweito Hii-seart und darf
vielleicht als Verschreibung au.s aramai.schem KjHOT (arab.
etwa -upo, erklart werden, obwohl aus den Gleichnngen
und (12 = gefolgert werden miisste, dass und (^'^)
mit einander identisch seien. So werden auch in der Jlisna die zwei
Hirsearten jnn und nebeneinander erwiihnt, und zwar bilden
Reis, die zwei Hirsearten und Sesam eine besondere Kategorie von
Kulturpflanzen zwischen Getreide und Hiilsenfriichten (Low, Pfl.
102 ; Flora d. Juden, i, 738), wahrend in unserem Falle D und ^•*00
offenbar infolge eine.s Irrtums in die Reihe der Hulsenfrlichte
hineingeraten sind. Tin. Frah. 1 Pahl., iv, 1, stehen die orter fiir Hirse
02, ydmru, arzan zwischen Weizen-Ger.ste und don Hiil.senfruchten.
Las.sen wir die zwei Worter und^ji^^ v eg, so bleibt in Gr. Bdh.
117, 2 f. als Reihe der Hul.senfriichte iibrig ;
wofiir der ind. Bdli. 64, 17 bietet :
Hievon ist A»|j^jj(g^/7neinn(Pazend Texts, od. Antia,49, 16 ;
von Justi, West (SBE., v, 101 " various kinds of pulse ) und Modi
(Pahl. Tran.sL, ii. 122 f.) fitl.schlich mit phi. yrmak identitiziert worden.
Es ist aber. wie .sich aus die.ser L’ntersuchung ergeben wird, nicht
zweifelhaft, dass das Paziindwort aus und zwar aus der durch
Weglassung des ersten ) entstandenen Form :^)y (die jiavuia
gelesen wurde) verschrieben ist. Denn es entspriclit dem des
JIITTELPERSTSCH lE\dK
549
Gr. Bdh. und dem des Frah. I Paid, in den Reihen der Hidsen-
friiclite.
Yon besonderer AViclitigkeit fiir unsere Untersuchung sind nun
noch die Stellen des Bundaldsn und des Zad.sparam. die von deni
Emporwacbsen einzelner Pflanzeu aus dem Mark, deni Blut, der
Lunge, den Hornorn. der Nase, dem Herzen de.s Urstieres liandeln.
Dieser Darstellung liegt der Gedanke zugrunde, dass zwiselien den
korperliclien Substanzen oder Organen de.s Stieres und den daraus
enstandenen Pflanzen eine nahere Bezieliung, zum Teile eine Westms-
gleicblieit (wie zwiselien Mark und 8esam oder zwiselien Blut und
Weill) besteht, und dass die Pflanze oder das pflanzliciie Produkt den
entspreclienden Korperteil der Gescliopfe, der Sesam das Mark, der
Yein das Blut, stiirkt. Koiner der drei Te.xte : ind. Bdli., 28, 5 fF.,
Gr. Bdh.. p. 93. 8 ff.. Zadsp.. 9. 1 ff. ist ganz korrekt. Am meisten i.st
wohl Zadsp. veriindert. der aber einigo fiir das Verstiindnis des Ganzen
wertvolle Zusiitze enthiilt. Hier kann nur der fiir unsere rntersucdiung
in Betraclit koniiuende Teil des Abscliiiittes behandclt werden.
Im Gr. Bdh., 93, 11 ff. heisst cs zunachst : liac niazr/ knujel (Ms.
olfenbar fiir da hier nicht am Platze wiire) nx
raS. ^vat-ic iHutfiP [//us?] .. aus dem Mark (entstand) der 8esam.
weil dieser das M'esen des (Markes hat : er i.st selbst auch ein iMark “.
Darauf folgt : hic uruv line trulk .. aus dem Horn
(entstand) aus der Nase (d.i. reuok) “. Im. ind. Bdh.,
28. 8 ff. ist offenbar zu lesen : hue muzg (Ideogramm zu kun'jPt :
nicht ddw 7). gut gut hur ciiP kP jxt tnazg uiPhuidu .. aus dem iMark
(entstand) der Sesam. einzelweise jedes eine Ding, das im Mark seinen
Sitz hat “. Daran schlies.st sich an ; hoc ■•iriiv hoc vPitlk
Es i.st nicht zweifelhaft, dass dieses »rir fal.sclie Schreibung des
sonst ^y), ^)y geschriebenen Wortes ist und auf einer Ver-
weclislung niit (/nm/uw/A' 65. 9, iip. gnndand) .,Lauch''
beruht. Sclion die Schreibung ^yy an der entspreclienden Stelle
des Gr. Bdh. sowie der L’mstand. da.ss das ‘^■^yy des ind. Bdh. ganz
ebenso unmittelbar neben steht. wie ^|i). ^)y in dor
niichsten Xachbarschaft von erscheint.
550
B. GEIGER —
beweisen, dass Justi. West und Modi mit Unrecht als „ Lauch
verstanden haben. Dazu komnit nun iioch die Darstellung in Zadsp.,
9. 2 and 4 (in Avesta. Paid., and Anc. Pers. Studies in honour of
P. B. Sanjana, Ixxviii ; SBE.. v, 177 f.). Auf die AVorte pas yurtak
ul rust hernach ivuchs das Getreide empor “ folgt
jj. was wie eine Aufzahlung der Getreidearten aussieht.
dies aber schon deshalb nicht sein kann, weil .. Sesam “ nicht
eine Getreideart ist. Uberdies wird durch das dann folgende ceyou
(= kuujet) tnazcj-cihrak (oder, ohne ceydn, (= kunjet) mazg-
cihrak\lli] rc7S), yvat mas hast awzagenak 1 mazg (.. denn der Sesam
hat das AVesen des Markes ; er i.st selbst in hoheni Masse ein Mehrer
des Marke.s “) vorau.sgesetzt, da.ss es vorhcr, wie in den Fassungen
des Bdh.. geheissen hat ,.aus dem Mark ist der Sesam (entstanden) “ ;
und ich halte es fiir moglich, dass eine
Yerschreibung daraus darstellt. Auf keinen Fall darf mit
West als gedeutet werden. Weit eher diirfte man eine Verschrei-
bung aus girgir^ annehmen, wenn nicht der von uns rekonstruierte
Zusammenhang der Stelle dagegen spriiche. Aber auch
sind nicht Glieder einer Reihe aufgezahlter Getreidearten, obwohl
an und fiir sich mit West als Yerschreibung aus oder
(Hirsearten) gedeutet werden konnte. Ich zweifle nicht daran, dass
aus (srur) [^] „ aus den Hbrnern (entstand)
verschrieben ist und irgendwie an diese Stelle ver.schlagen
wurde. Die.se Annahme wird noch durch den Umstand erhiirtet.
dass dieses dasganzahnlichindenzweiBdh.-Fa.ssungen
unmittelbar vor ^)y ^ (d.i. hac venlk venok) steht.
in unserem Zadsp. -Text sonst nicht vorkommt. Der Text ist also in
Lnordnung geraten. Urspriinglich ist offenbar auch in ihm zuerst
von der Entstehung des Sesams aus dem Mark, dann von der des
^ Vgl. oben die Stelle Gr. Bdh. 117, 2. Merkwtirdig ist, dass auoh in der
entsprerhenden Stelle Gr. Bdh. 93, II (vgl. oben) dieses erscheint, und
zwar in dem ganz unmoglichen
MITTELPERSISCH VEXoK
551
aus den Hornern. dann — und zwar v o r Blut und Wein — von
der Entstehung des aus der Nase die Rede gewesen. Von
handelt nun der 4. Paragrapli des Kapitels, den West zum Teil falsch
transkribiert und iibersetzt hat. Der Text lautet : ^ .
,Ois.
JSiiy) . Das hat West iibersetzt : ..... from the nose is the pulse {nidijs
or mdsah) which is called donak, and was a variety of sesame (siunagu).
[and it is] for other noses. “ Er hat also das er in der Te.xt-
ausgabe mats transkribiert hatte. dem np. mas gleichgesetzt. das aber
nicht die allgemeine Bedeutung .. pulse “ hat. Es ware nur m o g 1 i c h
dass uifl.s hier nicht .. Bohiie sondern — wie z.B. im Kurdischen —
etwa ., Linse “ bedeutet. Aber auch die Sclireibung
besonders das Schriftbild in dem Ms. K. .35 (vgl. Codices Avest. et
Pahl. Bibl. Univ. Hafniensis. vol. iv, fol. 241 v.. letzto Zeile). sprieht
gegen die Identitat mit mas. Ich hahe daran gedacht. dass
aram. meinen konnte. das aus aram. Xi.tsn .. Kichererbse '■
verschrieben wiire. Das ist moglich, aber unsicher. Micht zweifelhaft
aber ist, dass mit )^)y, "'ie das aus der Xase des Urstieres entstandene
genannt wird, die Hlilsenfrucht ^^y) gemeint ist. In den drei
folgenden M'orten ist denn auch noch ausdriicklich gesagt : )^)y)
(venok) ist sein Name (Qtl’ = ndm-as). Die letzten drei Worte
_ju^ bedeuten offenbar ,. [das )^)y)- das aus der Nase des
Urstieres entstanden ist.] ist auch wieder fiir die Nase [heilsam] “.
wie ja auch nachher von der aus der Lunge entstandenen Raute
(spand) gesagt wird, dass sie fiir (r«S) die Lungenkrankheit (heilsam) sei.
Wir geben nun die behandelten Textstellen in iibersichtlicher
Anordnung :
Frah. I Phi., 4, 2 ;
Gr. Bdh., 117, 2 : ^yi)i/\ ^y)
ind. Bdh., 64. 17 : IHJiT
Gr. Bdh., 93. 11 : ^yy (liac vettlk) {hoc sruv)
ind. Bdh., 28. 9 : (hacvenlk) (hacsrfir)
Zadsp.. 9, 2. 4 : • . • y^yy (hacvenlk) (hacsruv)
B. GEIGER
552
Aus dieser Ubersicht ergibt sich, dass usw. ebenso wie
usw. nur verschiedene Schreibungen e i n e r urspriinglichen
Form sind, und dass beide gleich naxvat (., Kichererbse “) Xamen von
Hiilsenfruchten sind. Und zwar ist es am wahrscheinlichsten, dass
und seine Xebenformen niit np. nilju wohl nur
Verscbreibung daraus), angeblich auch nilzii, identisch sind, das.
und zwar aucli ini Hindi, die L i n s e bezeiclinen soli. Es ware immerhin
m 6 g 1 i c h. dass die Form ^P- ^>^usang. musanj gehort,
die nach dem Burhan und dem Farh. i Su'url Xamen einer kleinen
Erbsenart sein sollen. Und konnte man zu dem ebendort als
Xame einer linsenahnliclicn Fruclifc angefiihrten np. 7nuiu stellen.
Doch ist die 'Walirsclieinlichkeit, dass es sich nur uni Verschreibungen
handelt. grosser. Fiir iMcilk- und inlcilg selbst kommt
naturlidi keine andere Erklilrung als die duxch np. inljii (so schon
Pahl.-Paz. Gloss, und Justi, Bdh., 243 sub .. “) in Betraclit.
Die anderen, von Junker, Frah. 115 versuchten Deutungen sind ganz
unmoglich.
Fiir und seine Xebenformen wird die Riclitigkeit der Lesung
venok erwiesen durcli das bezeichnender Weise nur im
XayinI vorkoinmende unok ,.Linse“ (0. Mann-K. Hadank,
Kurd. pers. Forsch., ni, i, 158; bei A. Querry in Mem. de la
soc. de lingu., 9 (1896), 116: vinouk, ritiik
lentille “). Dadurch werden aucli die traditionellen Lesungen
v7)/fiy, vlnuij, vl»u, r7//7 wenigstens zu einem Teile bestiitigt.
Es ist jedocli gar nicht siclier, da.ss .. Liiise“ die urspriingliche Bedeu-
tung des Wortes gewesen i.st. Ja, die Annalinie. dass liier eine
Bedeutungsverscliiebung stattgefunden hat, wie sie bei Pflanzennamen
ofter konstatiert werden kann,* wird sogar durch die folgende Erwiigung
wahrscheinlich gemaclit. Es i.st kaum noch feststellbar, oh die in
den oben erwiihnten Textstellen vorliegende Verbindung von venok
mit venlk .. Xase “ auf der Kenntnis einer sachlichen Bezieliung der
Hiilscnfrucht zur Xase oder auf einer willkiirlichen Zusammenstellung
zweier ahnlicher Worter beruht. Auf jeden Fall liisst sich jedoch die
Existenz eines sachlichen Zusarnmenhanges erweisen. Der hebriiische
^ Audi bei V. Hehn. Kultiirpfi u Haustiere ® 221 wird darauf liingewie^-en, da^s
die emzelnen Gattungen der Hiilsenfruchte sprachlich selbst in jungcren Epoehen
nitht sdiarf unterschieden werden, und dass sich, manchmal sogar in einera Wort die
drei Bedeutungen Erbsc, Bohne, Linse vereinigen.
MITTELPERSISCH VENOK 553
Name der Kicliererbse |l£K „ <116 kleine Nase “ stammt daher, dass
das schmalere Ende ihres Samens die Form einer Nase oder eines
Schnabels hat, wie man aus der Abbildung bei I. Low, Flora d. Juden.
ii, 438, ersieht. Auch der syrische Name der Platterbse (Lathyrus)
bedeutet „ kleine Nase“ (von X31t3‘in)- Er ist von der
Kichererbse, die im ‘Irak sogar noch geheissen hat, auf die
Platterbse iibertragen worden (Low, l.c., 427 und 438 f.). Wenn der bei
dem syrischen Lexikographen Bar ‘All iiberlieferte zweite, im ‘Irak
gebrauchte Name der Kichererbse mit Low, l.c., 438 zu lesen ist
und ,, gehbrnt “ (nach der Spitze des Samens) bedeutet, kbnnte diese
Bezeichnung der obenerwahnten Vorstellung zugrunde liegen, dass aus
dem Horn des Urstieres die Hiilsenfrucht mlcuk hervorgegangen
ist, die urspriinglich vielleicht auch nur die Kichererbse gewesen
ist. Man wird demnach annehmen diirfen, dass das Phi. -Wort venok
(mit -ok als Diminutiv'suffix) Ubersetzung eines aramiiischen
Namens der Kichererbse, wie ist. Es ist allerdings
auch denkbar, dass eine unabhangige Bildung nach der Art von
hebr. und syr. XiSuJHn vorliegt. Low, l.c., 427, hat darauf
hingewiesen, dass die Kichererbse magyarisch bagolyborso
,, Eulenerbse “ heisst nach den am Ende wie ein Eulenschnabel
umgebogenen Schoten. Mit Unrecht hat also Horn im Grundr.
d. iran. Phil., 1/2, 49 dem Nayinl-AVort die Bedeutung „Linse des
Augos (-bin ,sehend‘)“ beigelegt.^
^ Wahrend des Druckcs hat H. W. Bailey mich darauf aufmerksam gemacht,
dass er schon in London Philol. Soc. Transact., 1935. 6i, das von mir behandelte
Wort venule gelesen und zu dem NayinI Wort gestellt hat.
VOL. vni. PARTS 2 AXD 3.
36
I
Singhalesische Etymologien
By Wilhelm Geiger
"jP^ASS das Singhalesisclie unter die ,, Modern Indo- Aryan
Vernaculars “ einzureihen ist, wird heute woU von niemand
rnehr ernstlich bestritten. Im Liiif/uistic Survey of India, i, 1, p. 145,
ist diese Tatsache auch von dem Grossmeister der indiscben Sprachen-
kunde, Sir George A. Grierson, anerkannt vorden, obwobl in dem
Werke selbst das Singhalesische keine Beriicksichtigung finden
konnte. Moge es mir gestattet sein, einige ganz bescheidene Nachtrage
zu meiner vor nunmebr 38 Jahren erscbienenen Etjunologie des
Singbalesischen, die den Ausgangspunkt und die Grundlage fiir meine
Studien iiber diese Sprache bildete, dem allverehrten Jubilar zur
Feier seines 85. Geburtstages in aufrichtiger Dankbarkeit zuzueignen.
1) Das Suff. di. -hi (nom. sg. -Ua < *-Uija, *-hn'a) — altem -lla-ka
ist im Singh, noch einigermassen lebendig. Ich verweise auf hhili,
°lla ,, Achselgrube “ neben kiisa (Sk. kaksa). jndali, °della
,, Rasenstiick “ neben pida (Sk. pinda), vardli, °Ua .. leichter Regen “
(Sk. vdri). Vielleicht gelangen wir von hier aus zu einer Deutung des
Wortes duvili. °Ua ., Staub Es ist anzuschliessen an P.. Sk. dhupa
und bedeutet zuniichst ,, kleiner Rauch “. Semasiologisch darf auf
mihidum ,, Nebel wtl. ,, Erdrauch ‘‘ (Sk. main + dhuma) verwiesen
werden. Auch auf dhuma selbst konnte zuriickgegangen verden, da
V zuiveilcn an Stelle von m tritt, wie in navamu neben namanu ., .sich
beugen “ (Sk. namati).
2) Wie V fiir urspriingliches m, so kann auch umgekehrt m fiir v
eintreten. So in nama ,, neun “ neben nava. Dies gilt auch fiir
amunannvd ,, einfiideln “ (Clough ungcnau amunanavd), das = P.
dvundti, °ndti, Sk. dvrndti ist mit gleicher Bedeutung. Ein ger. avund
wird von Clough und Carter angefuhrt.
3) Es lasst sich, wie ich glaube, wahrscheinlich machen, dass in
der Proto-Singhalesi.schen Periode (4. bis 8. Jh.) die friihere Accentua-
tion des Sanskrit und des Mittelindischen sich iinderte. Es kam die
Tendenz auf, den Ton auf den Wortanfang zuriick zu ziehen und zwar
zuniichst auf die erste Silbe nur dann, wonn sie .schwer war : — — X .
Auf die Wichtigkeit der Unterscheidung von leichten und schweren
Silben fiir das Verstandnis des singh. Vocalismus zuerst hingewiesen
zu haben ist Helmer Smith’s Yerdienst, War die erste Silbe leicht.
556
\V. GEIGER —
die zweite schwer, so trug letztere den Accent : ^ — X . Waren die
beiden’ ersten Silben leicht, so scheint der Accent geteilt worden zu
sein, und zwar so, dass die starkere Betonung auf der zweiten Silbe
lag : ^ X . Nun beobachten wir baufig, dass im zweiten und
dritten Fall der Vocal a der ersten Silbe zu i, u verwandelt wird :
piyes ,, Gegend “ < P. padesa, siyan „ Dach “ < chadana, siyalu
„ all “ sctJi-cdci , dixt'd/H ,, Uufte ydyJictTia , HAj/vavcb ,, Stadt
< nagara. Stellen wir nun dagegen Formen wie sayuru „ Ocean “
< sdgara, mayil ,, Oheim “ < mdtula, pavuru ,, Mauer “ < pdkdm,
so ergibt sich, dass i, u in der ersten Silbe der Worter piyes bis
nuvara Schwachung des a-Vocals sein muss, fiir die kaum ein anderer
Grand gefunden werden kann, als die starkere Betonung der folgenden
Silbe. Wir haben also padisa, ckdddna, nagara nsw. vorauszusetzen,
aber sdgara, mdtula, pdkdra. Ein kurzer anlautender Vocal konnte
elidiert werden vor der betonten zweiten Silbe. Der Fall liegt vor in
lav „ Wildnis “ (nom. sg. Idv-a) durch *d0v aus P. atavf. Die Prae-
positionen «««-, upa- erscheinen als nu-, va- (durcb *dnu-, *uvd-).
So in vahan „ Sandale “ = P. updhand (zuerst wohl bei Ed. Muller).
Ich fiige dazu nuvd „ Saum am priesterlichen Gewand “ (nom. sg.
nuvd-va) = P. anuvdla, das wohl mit Stede von Wz. vd „ weben “
abgeleitet werden muss.
Der letzte Schritt zur grundsatzlichen Betonung der ersten
Wortsilbe, wie sie fiir die moderne Sprache giltig ist, mag in der
spateren Proto-Singhalesischen Zeit vollzogen worden sein.
4) Zum Schluss noch einige Einzeletymologien : 1) andinavd,
}iand° „ (ein Kleid) anziehen “ zu Wz. da „ bindeu “ + sam ; aMana
(°na-ya) ,, Unterkleid (eines Priesters) “ — P. sanddna. — 2) allas-a
„ Geschenk, Bestechung “ < *at-lasa, P. hattha + lanca. — 3) kakul-a
„ Fuss, Bein “ = P. kankala, Sk. kaiikdla „ Gerippe (cf. Turner,
Nepali Lang., s.v. kangdi) mit einem iihnlichen Bedeutungsubergang
wie bei dem deutscben Wort bein. — 4) kuldra ,, Balkon “, Dhatnma-
padattliakatJid-Guta-pada, ed. Sir D. B. Jayatilaka 34®, 215*^ =
P., Sk. kvtdgdra, aber Singh, kotdra = P. koUlidgdra (Turner, s.v.
kotho). — 5) kolanavd ,, pressen, gliitten, platt macben “, DhpAGp.,
25^^ = P. (d-, paccd-)koteti ; aber kotanavd „ zerbauen “ = P.
kotteli. — 6) pild-va ,, Milz “ = Pk. pilikd, Sk. plihan (Turner, s.v.
phiyo ; das Singh, steht bier dem Oriya, HindostanI, Panjabi naher
als dem Marathi). — 7) muhu, mu ,, reif “ = P. mudu, Sk. mrdu. —
8) mohot-a ,, Augenblick, Moment “ = Pk., P. muhutta, Sk. muhurta,
ein Beispiel fiir die Neigung zu offener Aussprache des Vocals u. —
SINGHALESISCHE ETYMOLOGIEN
557
9) Das Wort vahan fiihrt uns auf das honorific vahanse. Hier ist se =
P., Sk. chdya in dem Siime von ,, Abbild “ zu fassen ; vahanse ist
also Abbild der Sandale, d.i. Fusspur. Man sagt in respektvoller
Rede svdmm-vahanse statt kurzweg ,, Herr, Herrscher Bekanntlich
spricht man im Sanskrit ganz ebenso von den pdddh einer hervorra-
genden Person, man verehrt beispielsweise guroh pdddn. Da nun aber
hier pdda nie im Dual gebraucht wird, so bedeutet das Wort wohl
nicht ,, Fusse “, sondern auch ,, Fusspuren Beide Ausdrucksweisen
sind vereinigt in einer Inschrift Vijayabahu’s ii, 1186-7, wenn wir
Epigraphia Zeylanica, ii, p. 183^® den Text (nach Magd. Ggr.) so zu
trennen haben ; Simhabdhu-rajapd-vahanse. Das ware Pah S.-rdja-
pdda-updhand-chdyd. Die urspriingliche Bedeutung von vahanse
hat man iibrigens nicht mehr gefiihlt. Man kann es auch von vereh-
rungswurdigen Gegenstanden gebrauchen, wie z.B. dhdtiinvahanse,
etwa ,, heilige ReUquie Von vahanse kommen wir endlich zu einer
Erklarung des pron. der 3. p. unndnse das von Respektspersonen
gebraucht wird. Es steht fiir *unnahanse < *unvakanse. Hier ist
un (wie oben svdrnin, dhdtun) C. obi. pi. (maiestaticus) und es bedeutet
unndnse etwa ,, Seiner Ehrwiirden — 10) sumhulu-va ,, Rolle,
Bolster, Hiille, Membrane “ = P. cuntbata. — 11) kaputu-vd und 12)
kamdu-vd „ Krahe *• mogen den Beschluss bilden. Ich fuhre die
beiden Worter zuriick auf M.Ind. *kapputtha. bzw. *kawuddha
< kad + p°, bzw. v°, Sk. *kat-pusta, bzw. *kad-vrddha „ schlecht
(mangelhaft, elend) ernahrt, bzw. grossgezogen.“
The change of a to e in the Indo-Aryan Loan
Words of Malay alam
By K. Goda Yaema
[Abbreviations of languages in this article are the same as those employed in
Professor Turner’s Nepali Dictionary. Forms without any abbreviation are
Malayalam words.]
T he pronunciation of a coming after certain initial consonants as
e, is the most important of the vowel changes met with in the
Indo-Aryan loanwords of Malayalam, and also perhaps the most
interesting from a phonological point of view.
The first to note and record the change was Gundert, He observed
that a occurring after the voiced sounds g, j, d, d, y, and r is heard as
e in the pronunciation of the Malayalis ^ (A Grammar of the Malayalam
Language. 2nd ed., p. 8). Gunderfs observation is obviously imperfect
inasmuch as the change is regularly manifested after h and I also,
over and above the sounds specified by him. Professor Rajarajavarma
remarks : " a is the heaviest of all vowels in Malayalam and as such
it has undergone corruption in various words. The corrupt
pronunciation of a in Malayalam is very much like e." E.x. Sk. gandJia-
gendham, Sk. jana- jenam, Sk. dambha-^ dembhu, Sk. dayd deya, Sk.
bandhu- bendhu, Sk. raid- fevi, Sk. lajjd lejja. The words quoted above
show that in Sanskrit loans a following unaspirated voiced stops
{mrdus) which remain voiced in Malayalam, the sonant y, and the
liquids r and I is pronounced e though -m-itten a ’’ [Kefalapdninlyam,
4th ed.. pp. 86-7).
Xeither Gundert nor Rajarajavarma has explained the change.
What I propose to do in this article is to discuss in full the scope of
the change under consideration, and also to attempt an explanation
of the tendency so peculiarly marked in the pronunciation of the
Malayalis.
It will be seen from the examples cited by Professor Rajarajavarma
that the sounds after which a manifests a tendency to be pronounced
as c are all voiced consonants occurring initially, which latter are
totally absent in native Malayalam words. The change in question
^ Malayalam speakers who are also natives of Kerala.
^ damhha is Prakrit, the corresjpondmg JSaiisknt word being dambha-.
560
K. GODA VABMA —
is seen not only in Sanskrit loans, but also in other Indo-Aryan words
that have found their way into the Malayalam vocabulary. The
following examples from Middle Indian and New Indian will illustrate
the point. Middle Indian : Pk. jamula- jevili “ double cloth, cloth
in general ” ; Pk. ^mbha- dembhii “ pride ” ; Pk. ratthia- rettiydr
“ one belonging to a class of Telugu Sudras ; Pa.Pk. lakkha- lekku
“ aim ” ; Pk. laddi letti ‘‘ dung of animals New Indian ; H. jati
jeti “ musical rhythm ” ; H. dabbd deppa “ a tin vessel for holding
oil ” ; H. bardl bedayi “ self praise, magnifying ” ; H. lattu lettu
“ a child’s top ” ; M. dambdi dempriti “ a copper coin of small value ” ;
M. dasrd desra “ a festival held in honour of Durga ” ; M. lakhotd
“ closed letter ”, lekkottu “ envelope ” ; H.M. dabbd deppi “ a small
box ” ; H.M. bangld behkaldvu “ a bungalow ” ; H.M. Idgotl lenkdtti ;
H. laddu, M. laddu leddu “ a globular sweetmeat ”.
Indo-Aryan a preceded by palatal stops (both voiced and voiceless)
is also found represented by e in a fairly large number of loanwords.
The change which is evidently due to the position of the tongue
required for the production of palatals colouring the vowel following,
is often met with in the colloquial dialects in native Malayalam words
showing a in the literary dialect. Ex. cakini, cekini “ layers of fibre
enclosing the flesh of jack fruit ” ; cakifi, cekiri “ coco-nut fibre ” ;
caviitu, cevittu kick ” ; cali, celi “ mud ” ; canvu, cefivu “ slope ”.
The following forms with e in hterary Malayajam, a few of them
showing alternatives with a and e, may be regarded as loans from the
popular- dialect. Sk. candra- “ gold ”, centram “ golden neck
ornament ” ; Sk. campaka-, cempakam “ michelia champaca ” ; Sk.
earn- cafuvam, cefuvam “ cooking pot ” ; Sk. jhatiti jhetiti " soon ;
Sk. jhampada- cempata “ a mode of beating time ” ; Pa. Pk. camma-
cemmdn “ a worker in leather ” ; Pa. Pk. cakka- cakku, cekku " oil
press ”.
AVith the above may be compared Sanskrit loanwords in Kanarese
showing a similar change. Ka. candi, cendi “ a passionate, violent
female ” ; Ka. canda, cenda “ lovely, charming ” ; Ka. candira,
cendira ” moon " ; Ka. carce, cerce “ smearing the body with sandal ”.
It remains to suggest an explanation for the phonetic development
detailed above. Malavalam a is not the equivalent of Hindi or Marathi
a. In pronunciation it takes its place between numbers four and five
of the cardinal vowels, more inclined towards number four than five,
and evinces a tendency tow'ards greater closeness under favourable
circumstances. In the case of voiced consonants an occasion for the
CHANGES IN INBO-AKYAN LOANWOEDS OF MALAYALAM 561
natural tendency to manifest itself is provided by the lack of muscular
intensity characterizing their pronunciation, a preceded by palatals
also has changed to e in colloquial Malayalam in a fairly large number
of words (cf. the previous paragraphs). It may be observed that the
influence of the position of the tongue required for the production of
the palatal consonants is, like the lack of muscular intensity noted
above, a favourable factor conditioning the change. The above change
is not observable when a follows a voiceless stop or voiced aspirate
(palatals excluded), the reason being the resistance offered by the
greater muscular tension involved in their pronunciation, ex. Sk.
kaksd kacca hem of the garment tucked into the waistband, cloth
in general” ; Sk. cakra- cakram “wheel, a coin valued 1-28 of a
rupee ” ; Sk. taksan- taccan “ sawer of wood " ; Sk. paksa- paksarn
“partiality”; Sk. khadga- kkalgam “sword”; Sk. chanda-
■■ appearance, look, shape ”, cantam “ beauty” ; Sk. phaJaka- palaka
“ plank ” ; Sk. ghanta ghan^ “ bell ” ; Sk. dkana- dhanam “ wealth,
riches”; Sk. bhahgi- “way of dressing, fashion, toilet”, bhangi
“ beauty ”. An interesting parallel of a phonetic change affecting
only voiced consonants is provided by the cerebralization of dentals
in Sindhi. The surd stops and aspirates which are pronounced with
greater energy have remained dentals in Sindhi, while d alone has been
cerebralized. (E. L. Turner, JRAS., 1924, p. 583.)
Voiceless stops of certain loanwords in Malayalam, going back to
voiced stops foUow'ed by a in Indo-Aryan, are pronounced with e in
colloquial Malayalam. Ex. Sk. gati~ “ movement, procedme ”, keti
“ means, way ” ; Sk. garu^- kerutan “ Visnu’s bird ” ; Sk. garbha-
kerpam “ pregnancy ” ; Sk. garva- keruvu “ pride ” ; Sk. danda-,
OMal. tentam “ fine, punishment, trouble " ; Pa.Pk. ganthi- kenti
“ inflammation of the scrotum ”.
In Tamil e is not only pronounced after voiceless stops repre-
senting Sanskrit voiced stops, but also written. Ex. Sk. gaja- Tam.
kecarn ; Sk. gandha- Tam. kentam “ odom ” ; Sk. gambhira- Tam.
kempirarn “depth, profundity”; Sk. darsana- Tam. tericanam ]
Sk. bandhu- Tam. pentu “ relation, friend ” ; Sk. bala- Tam. pelam
“ strength ” ; Sk. bandha- Tam. pentam “ tie, bondage ”.
The foregoing e.xamples would suggest that Malayalam and Tamil
had a period when Sanskrit initial voiced stops were pronounced voiced
in the loanwords. The appearance of o in colloquial Malayalam where
we would natmally expect e as in kaiicdcii “ hemp " (Sk. gaiijd),
tantu “ stick, pole ” (Sk. danda-), pamparam “ spinning top ” (Sk.
562 CHANGES IN INDO-ARA'AN LOANWORDS OP MALAA'ALAM
hambhara- " bee ’), tavi ladle’’ (Pk^.daiTl) mav be regarded as the
result of the attempt on the part of the learned to maintain, as far as
possible, the pronunciation of Sanskrit words as written, influencing
the colloquial speech. In this connection mav be contrasted the
retention of e after k by Malaya jam, both in the colloquial and the
literary dialects, in native DraAudian words where Kanarese shows
ga- < Primitive Dravidian ga-. (For the general loss of voice in the
initial voiced stops of Tamil and Malayalam see Pre- Aryan and Pre-
Dravidian, p. 48, “'Sanskrit and Dravidian,’’ by Jules Bloch.) Ex.
kettu in anakkettu “ dam ", Kan. gattu ; kettii ” bundle ", Kan.
gaUe ; kettu in kettu-tati “ beard about the chin ’’, Kan. gadda ; kentan
" large ", Kan. gandu ; keta “ trembling palpitation ”, Kan. gada ;
kenakku " to roar, sound harsh ’’, Kan. gane.
In the above instances, unlike the case of Sanskrit loans, there is
nothing to tell the Malayajam speaker that the e represents earlier
a, and in the absence of any force influencing e towards being
pronounced a, e has remained unchanged both in literary and colloquial
Malayalam.
In the pronunciation of y, r, and I both the factors conditioning the
change, namely the lack of muscular tension and the palatal position
of the tongue,^ are present. Hence a following these sounds also shows
a regular change to e. \Vith this may be contrasted a remaining
unchanged after v, which latter is the only voiced sound occurring
initially in native Malayalam words. Ex. Sk. vatu- vatani " rope " ;
Sk. varna- varnani " colour " ; Sk. valli- vaUi creeper ’’ ; Sk.
varddi mrddi a mu.sical tone " ; Sk. vana- vaham " forest " ; Sk.
vayas- rayassu age ’’. The labial position of v may also have been
responsible to a certain extend for resisting the change.
In conclusion, then, we .see that the change of a to e in the Indo-
Aryan borrowings of Malayalam is conditioned by the lack of muscular
intensity of the consonant preceding a or the palatal position of the
consonant with which a is associated.
^ In Malayalam r and / are pronouneetl with the tongue ])Iafcd in a ])alatal jiu.'sition.
It may be noted that there are t’W'o kinds of rm Malayalam, one palatal (r) and the other
alveolar (r), and that the alveolar r occurs only very rarely in the Indo* Aryan
borrowings.
Observations on Middle Indian Morphology
By Louis H. Gray
TNVESTIGATION of Middle Indian morphology from the strictly
linguistic point of view can fairly be said to have been made thus
far only by Jules Bloch, notably in his L' Indo-aryen du Veda aux
temps modernes (Paris, 1934). However valuable as descriptive
grammars and as collections of material the Grammatik der Prakrit-
Sprachen of Richard Pischel (Strasbourg, 1900) and the Pali Literatur
und Sprache of Wilhelm Geiger (Strasbourg, 1916) undoubtedly are,
both works are far from linguistic in purpose. In Bloch's masterly
survey of the history of Indian linguistic development from Vedic
through Sanskrit and iliddle Indian to Modern Indian, on the other
hand, embracing phonology, morphology, and sentence-structure,
it was scarcely possible, in view of the mass of material, for him to
discuss every detail. It is my purpose, then, as a comparative linguist,
to consider in the following pages certain phenomena in Middle Indian
which seem to merit further study, omitting on principle all that
appears already to have been satisfactorily e.vplained, such as the
pronouns (cf. Bloch, pp. 145-7). Speaking in very general terms.
Middle Indian would seem to present a mixture of forms common to
Vedic and Sanskrit, a number of survivals to be paralleled only in
Vedic or Iranian, and a considerable amount of contamination of
formations whose functions were, at least approximately, identical.
I. XoUNS
The nominative singular of -u-stems ends in -6 in the majority of
Prakrits and in Pali, but in -e in MagadhI and ArdhamagadhI, and in
-u in Apabhraihsa (Sanskrit putrds : Mg, AMg putte : A puttu). The
two latter are reflected in Old Siriihalese {piite) and perhaps in Old
Bengali {kumhhlre " crocodile " ; Skt. kumbhiras), and in Old
Rajasthani and Sindhi (paw = Skt. pddas "foot’’, deh'‘ ~ desds
country "’) respectively (Bloch, pp. 167, 168) : while -e is at least
superficially found in such Gafla Avesta forms as vai\> = Skt. vdcas
“ speech ” (C. Bartholomae, in Grundriss der iranischen Philologie,
i, Strasbourg, 1901, pp. 154, 179). The Middle Indian -e is explained
564
L. H. GKAY —
by E. 0. Fraake {Pali mid Sanskrit, Strasbourg, 1902, pp. 104, 116,
128) as for -d-i, a view which scarcely deserves even mention. The
question arises, rather, whether a purely phonological explanation
may not be in order. The Skt. -d for -as, -ah has long seemed to me to
be, not a true [o;], but a mere fagon d’ecrire for a very different
phoneme. In Skt., -as < *-os (e.g. dsvas < *ekuos) is preserved before
voiceless occlusives {dsvas pdtati “ the horse falls ”), but is written
-0 before voiced occlusives {dsvd dhdvati “ the horse runs ”). Now
dsvo dhdvati is obviously for *dsvaz dhdvati ; but [z] not being a Skt.
phoneme, it vanishes with compensatory lengthening of the preceding
vowel. In other words, *az > *« > * [o;], so that the real
pronunciation of dsvo dhdvati was probably something like [asvo;
d‘a:vati]. Beside this, however, there would seem to have been the more
regular development of *az to [a:], i.e. [asva: d‘a:vati] beside [asvo:
d‘a:vati]. Between these two there would appear to have been inter-
mediate stages in which the relevant vowels were obscured, and one
may accordingly suggest that the type of -putte is a “ dulling ” of
[putta:] < *[putra;], and puttu of [putto;] < [putroij.
Developments analogous to those here suggested may be seen in
Anglo-Saxon stdn [sta:n] > iVEddle English stane, ston [sto:n]
> English stone [sto:n] > [stoun], but Scots stane [stern] (in Northern
English the “ rounding” of [a:] did not take place), north-east Norfolk
and east Suffolk [stAn] (cf. J. Wright, English Dialect Grammar,
Oxford, 1905, pp. 88-92).
In the ablative singular, Maharastrl, Jaina Maharastri, and AMg
puttdo (M, etc., also puttdu), Saurasenl, Jaina Sauraseni, Mg puttado
(JS also puttddu), PaisacI puttdto, puttatu appear to be contaminations
of putrdt and putrdtas, unless, as my pupil, Mr. AVilliam B. S. Smith,
suggests, they are for the ablative putrdt plus the genitive-ablative
-as < Indo-European *-es, *-os of consonantal stems (cf. the A
genitive singular puttahb, puttassu contaminated from putrdsya
and *putrds-as, Pischel, p. 251). The view of Pischel (p. 63)
that a is arbitrarily lengthened before -tas seems quite without
foundation.
M puttdhi, puttdhinito, and A puttahe, pultahu are more complex.
The form puttdhi seems to presuppose *puttdbhi, perhaps a contamina-
tion of *putr6t (Skt. putrdt) with an instrumental singular *putrcbhi
(cf. Greek deo-cfyL, Armenian mardo-v), this being aided by the
confusion between the ablative and the instrumental in the plural
in AMg and JM, where both cases are puttehim (sometimes
OBSERVATIONS ON MIDDLE INDIAN MORPHOLOGY
565
strengthened in the AMg ablative by the ablatival -tas, jputtehimlo).
The form puttdhi would seem to be identical in tj'pe with Skt. (SB)
lUtardhi “ from the north ”, daksindki “ from the south ” (Panini,
V, iii, 37 : dhi ca dure).
Elsewhere also, as is well known, the instrumental plural has
usurped the functions of the ablative, as in Greek, Latin, Old Irish,
and Gothic (Aukoi?, lupis, feraib, inilfam ; cf. also the dative-ablative
plural -m- in Slavic beside the instrumental singular -m-, e.g. Old
Church Slavic vlukmnu : vlukomi, and Greek ded<^t(p) as instrumental,
ablative, and locative alike, both singular and plural). The formative
-bhi- likewise appears in the dative singular of certain pronouns
(e.g. Skt. tubhyatn, GAv tna'byd, Latin sibi, Oscan sifei), in the
instrumental of all numbers (sing. ; Greek dedc^t, Arm. mardov ;
plur. : Ved. devebhis. Old Irish feraib ; dual : Skt. devebhydm,
GAv zastoibyd, Old Irish feraib), and in the dative-ablative plural
(Skt. devebhyas, GAv dddae'byd, Latin dedbus, Oscan luisarifs, Gaulish
vafxavoLKa^o). The base-meaning seems to have been “with”,
but — at least in the historic period — the formative was common to
several cases and to all three numbers (cf. Gray, in Language, viii
[1932], 191, 192-3, 194, 195, 197).
The A ablative type of puttahe is apparently derived from
*putrdsyds, i.e. it has the termination of the feminine genitive-ablative
pronoun, a like confusion of genders recurring in the AMg, JM dative
puttde < *putrdyai (Pischel, p. 249). The form puttahu, on the other
hand, seems to point to an original *putrdbkas, i.e. to a formative -bh-
(as in dedbus, luisarifs, vagavaiKapo). as contrasted with -bhi- (as
in devebhyas, dddae'byd), plus the genitive-ablative termination -as.
Derivation from *putrds- plus -as seems much less likely, and Pischel's
interpretation (p. 257) of the A ablative plural puttahu as equivalent
to the Skt. ablative dual putrdbhydm appears highly improbable.
The nominative plural masculine and neuter, represented by AMg
puttdo (masculine), M, AMg, JM, S, Mg phaldim, phaldi, phald'i
(A, M also phala'i) (neuter) likewise call for consideration. Pischel's
view (p. 254) that puttdo represents Skt. putrds plus the nominative
plural termination -as of consonantal stems is obviously incorrect,
nor can the Prakrit form be compared with the Vedic type putrdsas,
Av. aspduho. Old Persian bagdha (the latter probably borrowed from
Medic ; see Meillet-Benveniste, Grammaire du lieux-perse'-, Paris,
1931, pp. 10-11, 175) since h < s does not disappear in Prakrit (Pischel,
pp. 183, 184). The type of puttdo, like those of AMg, JM, S agglo.
566
L. H. GRAY —
AMg. JM vaiio (contrast Skt. agndyas, vaydvas), would seem to be
formed directly from the attested Prakrit nominatives plural putta,
aggl, vdu plus the plural termination -as of consonantal stems.
Similarly the feminine t}’pes of M, AMg, JM, S. Mg mdldo, devlo.
vahuo are from *nidldas. *devtas, *vad}iuas (contrast Pali mdldyas.
but deviyo, vadhuyo, as opposed to Skt. maids, devyhs [Yedic devts'\,
vndhvds), with the long vowel carried throughout, plus the con-
sonantal termination -as (cf. Bloch, Formation de la langue nmraihe,
Paris, 1920, p. 186).
The neuter tvpes of phaldim. dahlim. malmim. continued in the
Marathi -em (Bloch, pp. 81, 185 ; cf. R. L. Turner, JRAS 1915,
p. 27 : 1927, p. 230) and represented in Sanskrit by phdldni.
dddhlni, nuidhuni. have their long vowel from the t\’pes represented
by Yedic p/irtld, dddhi, mddh% (cf. Young Avesta xsadra, zaradustri,
fo'^ru — quantity of final vowel ambiguous). The final -irh, -i, -i
can come only from *-in{i), borrowed from the -f«-stems. Here one
would expect *-ln{i), i.e. *pJiald7m, etc. (cf. Skt. bdlini) ; but the -i-
is clearly due to analogy with the -i- of the Prakrit nominative and
accusative plural masculine of the type of aggino. seen also in the
instrumental singular aggind and in the genitive singular aggino
(cf. Skt. instrumental singular agnind, Pali aggind, genitive aggino.
ablative aggind ; cf also the short vowel in Av. vohungm, Old Persian
dakyioidni). Against this h\’pothesis on^ may fairly argue that final
-i is preserved in the Mg ablative singular piittdki (cf. A puttdhe),
in the M, etc., locative singular puftammi (cf. A puttahi) and
instrumental plural puttehi(m), puttehi, and in the A locative plmal
pnttahi ; and that the only loss of a final vowel recorded for Prakrit
noun-infle.xion is in the t}'pe of the A instrumental singular pidtem.
pvtti (cf. also Pischel, pp. 111-12). A solution may possibly be sought
in the fact that in stems in -n- (and -«/-), from which -dni — and.
consequently, -Ini. -uni — is borrowed (Debrunner-Wackernagel,
Aliindische Grarnmntik, iii, Gottingen, 19.30, 105, 160-2. 276-7), Av
shows forms in -an < *-dn beside tho.se in -dni < *-dno (e.g. GAv
ddntnm : afPndni). and this ending is borrowed in Av -a-stems (e.g.
G.Vv pnamm. YAv aino.syg.n : GAv akd. YAv xsadra. Old Persian
hamarand [Bartholoniae, pp. 22.5, 233 ; cf. also K. Brugmann.
Grundriss der vergleichenden Grammatik der indogcrmarnschen
Sprachnn^, Strasbourg, 1897-1916. ii. ii, 232]). In Prakrit phaldni.
phaldim, etc.. I see, accordingly, a survival of the Indo-Iranian
doublets *-ni and *-n.
OBSEKVATIOXS OX MIDDLE INDIAN MORPHOLOGY
567
11. Verbs and Verbals
In the present indicative active only the A t\’pes of the first singular
vattau, the first plural vattahu, the second plural vattahu. and the
third plural vattnhi are not self-evident. The t}’pe of vattau, whose
termination survives in Old Hindi, Braj, Old Gujarati, and
Chhatlsgarhl -an (Bloch, I ndo-aryen . pp. 248-9 ; cf. R. Hoernle,
Comparative Grammar of the Gandian Languages, London. 1880,
p. 335), is well explained by Pischel (p. 322) as for *vdrtakam ; but
his view that this is to be compared with Skt. ydnmki = ydmi (KB
xxvii, 1) and the forms pacataki, chindhaki ~ pdcati, chindha, etc.,
quoted by Skt. grammarians, seems improbable (cf. L. Renou,
Grammaire sanscrite, Paris, 1930, p. 247). In *vdrtakam I incline to
see, rather, the element -kam which appears in the nominative singular
of the first personal pronoun M ahaatii, JM ahayam, Asoka hakam,
Mg hag{g)e, A, Panjabi, Braj had, etc. = Skt. ahakdm < *eyhm-k6m
(cf. Pischel, p. 293 ; Bloch, pp. 145, 191, and Memoircs de la Societe
de Linguistique de Paris, xxiii [1929], 115). This element -kam is
of doubtful origin, but is possibly the formative which recurs in the
Indo-Iranian “ genitive plural '' (Skt. asmdka{m), yu.v)idka()H), Av
ahmdksm, yusmdkom). In any event, it would seem that the pronominal
ending -au has here influenced the verb (cf. also the Pali first singular
optative vatteyydhani < *vatteyyam aham, Asokan vattc 'ham, and
Pali first plural optative vatteyydmhe < *vatteyyd amhe [Geiger,
p. 75 : Bloch, locc. citt.]).
Whether the second singular vattahi is to be compared with Skt.
vdrtasi (Pischel, pp. 183, 322) or— as seems rather more likely —
with an imperative *vdrtadhi (Bloch, Indo-aryen. p. 247), it, in con-
junction with the second plural rattaha. plainly accounts for the
otherwise inexplicable h in the first and third plural vattahu, vattahi.
The form vattahu itself presupposes *vdrtathas. with -as instead of -a
(cf. Skt. vdrtatha) on the analogy of a first plural *vdrtahas (cf. Skt.
vdrtdmas ; that we here have a dual — cf. Skt. vdrtathas — seems
rather less likely).
In the first person plural one would expect * vattahu (cf. Skt.
vdrtdmas, general Prakrit vuttdmo), but, besides the shortening of
d by analogy with the rest of the present, *-u has become -u.
apparently under the influence of the singular vattau. In the third
plural vattahi, represented in Modern Indian by Old Gujar.ati ndcui
“ they dance ", Awadhl and Braj calai " tliey go ”, and Chhatlsgarhl
ghuchal'' they can move " (Bloch, pp. 248- 9 ; Turner. JRAS 1927,
568
L. U. GRAY
p. 238 regards the nasalization as representing the older termination-
nti). the nasalization, perhaps drawn from the first plural, is added
to the singular vaitai, so that the relation may be summarized
in the scheme : —
First sing. vatMu ; first plur. vattahu = third sing, vaitai :
third plur. vattahi (cf., on the whole question, Bloch, pp. 247-9).
In the middle, a single Pali form merits consideration here. The
first plural of the type of vattamhe (with the semantic distinction in
Pah between vatt- “ exist ”, etc., and vviit- “ be right ” we are not
concerned in this connection), cited only by grammarians, is clearly
for *vdrtastne, and can scarcely arise from syncope of *vattdmahe
= Skt. vdrtdmahe, despite Geiger, p. 107, and V. Henry, Precis de
grammaire pdlie, Paris, 1904, p. 80. Just as in the first plural active
of the tj’pe of avattamha (see below, p. 570), we evidently have a
formation analogous to — perhaps actually combined with— the
copula *amhe (cf. the active amhd — Skt. smds) with -e mechanically
substituted for -d as a “ primary ” middle sign on the analogy of
vattasi : vattase ; vattati : vattale ; vattanti ; vattante, the proceeding
being aided by the -e of the second plural, where it is fully justified
(cf. Pali vattavhe [attested only by grammarians] ; Skt. vdrtadhve,
and see Bloch, p. 232). The type of vaitamase beside vattamhe, which
recurs in the imperative and optative t3q)es vattamase, vattevmse,
shows merely another mechanical substitution of the “ primary ”
middle characteristic -e for the active -i (cf. the Vedic first plural
active vdrtdmasi) ; and the type valtamhase appears to be simply a
contamination of the two types vattamhe and vattamase (Geiger,
p. 107). The aorist middle type of avattamhase is probably of identical
origin (see below, p. 573).
The optative has been discussed by Pischel (pp. 325-9, and KZ
XXXV [1899], 142-3), H. Jacobi (ibid., p. 577), Geiger (pp. 110-11),
and especially Bloch (pp. 2-33-4, and “ Quelques desinences d’optatif
en moyen-indien epigraphique et litteraire ”, in MSLP xxiii, 107-120).
It falls into two types, represented by Prakrit vatte = Skt. vdrtet
(Type A) and I'attejjd < *vdrlegat, *vdrteyat{i) (T\pe B) respectively.
The first of these calls for little comment except for the tj'pes of the
Asokan third plural vattevd < *vdrtevur beside vatteyn (contrast Skt.
vdrteyiir), where I can account for v instead of y only on the very
unsatisfactory assumption of an arbitrarily sporadic interchange of
the two phonemes (cf., for other instances, Pischel, p. 177), and Asokan
vattavu, a contamination of the optative represented by Skt. vdrteyur
OBSERVATIOXS ON MIDDLE INDIAN MORPHOLOGY
•569
with a subjunctive *vdrtdyur, again with v fox y ; the general Prakrit
third plural type vatte < *vdrten ; the Pali first singular vatte
< *vdrtem ; and the first plural vattemu < *vdrtemu, where one
apparently has a contamination of the optative (cf. Skt. vdrtema)
with the imperative (cf. Skt. vdrlatii) (Bloch, MSLP xxiii, 118 ;
that -u is a Prakritism for -as [cf. AMg accemii, etc., and see Pischel,
p. 235 ; Geiger, p. 81] is less likely).
Type B is more complex. It would appear to be a contamination of
an analogical extension of the type of Skt. vdrteyam with the
subjunctive (PischeTs ^dew, that the contamination is of the thematic
and the athematic optatives, seems less hkely). Rather intricate in
detail, the type will perhaps most clearly be presented by the following
comparative table : —
Prakrit.
Pali.
Skt. optative.
Skt. subjunctive.
Singular 1 .
vattejja
\'ed. vdrtd
vattejjdmi
valteyydmi
*vdrteydmi
vatteyyam
vdrteyam
(cf. As. avatte ’ham)
vatteyydham
*vdrteyam ahdm
„ 2 .
vattejjdsi
vatteyydsi
*vdrteydsi
vdrtds{i)
vattejjasi
vatteyya
*vdrteyae(i)
3 .
vattejjd
vatteyydti
*vdrteydt{i)
vdrldt{i)
vattejja
vatteyya
*vdrteyat
Plural 1
vattejjdma
vaiteyydma
*vdrteydma
vdrtdma
•4
vatfejjdka
vatteyydtha
*vdrteydtha
vdrtdtha
vattejjaha
*vdrteyatha
„ 3 .
vattejjd
*vdrteydn
vdrtdn
vattejja
*vdrteyan
vatteyyu(m)
vdrteyar
It is possible, furthermore, that the “ injunctive ” of the type of
Skt. {d)vartam, {d)vartas, {d)varlat, (d)vartdma (but not [d]vartain),
{d)vartan, may' also have played some part in the formation. The
Prakrit first singular vattejjd seems to be a contamination of the Vedic
subjunctive vdrtd with the optative vdrteyam, as vattejjdmi (Pali
ratteyydmi) is of vdrteyam with the present indicative vdrtdmi ; and
the type of vattejja is probably by analogy with the third singular
vattejja (otherwise Pischel, pp. 326) ; for Pah vatteyydham, etc., see
above, p. 567 ; for -a < -as in the Pali second singular see Geiger,
pp. 73, 128, 129.
In the imperative active the -u characteristic of the third persons
{vattaii, S, Mg, Dhakki vatfadu : vattantu = Skt. vdrtatu ; vdrtantu)
has analogically influenced the first singular (recorded only by
grammarians) vattdmu, the second singular vatfasu, vattesii, and the
VOL. Vin. P.4 RTS 2 .4X0 3. 37
570
L. H. GRAY —
A second plural vattahu, vaUehii, this explanation (cf. Pischel, p. 331)
being much more plausible than the older view (repeated by Bloch,
hido-aryen, p. 249), which makes the termination a transfer from
the middle voice (Skt. vdrtasvn, Pali rattassu), in which event one would
expect a Prakrit *vaitassa or *valtassa (cf. Pischel, p. 218).
The first singular vattdmu < ^vartdm-u corresponds to the formation
of Skt. vdrtdmi, while vaUamu < *vdrtmn-u, outwardly similar to the
Pah first singular indicative vattmh beside vattdnii, seems to be, in
reality, an old “ injunctive ” (cf. Skt. dvartam), just as the A second
singular vaUu appears to be for *vdrtas (cf. Skt. dvartas). The A second
plural vattahu may either be the indicative used as an imperative (as
is true also of the second plural vattaha and of the third plural vattahi)
or may have arisen from *vdrtathu, with -th- instead of -t- through
analogy with the indicative (cf. S, Mg vattadha, vaUedha, general
Prakrit vattaha, Pah vattatha contrasted with Skt. vdrtata), exactly
as the second singular imperative vat^su < *vdrtas-u corresponds
mechanically to the second singular indicative vattasi — Skt. vdrtasi ;
i.e. we have here a contamination of the imperative and the “ injunctive ”
The A second singular vat^hi would seem to be for *vdrtadhi, with
-hi < -dhi carried over from the “ athematic ” conjugation ; and the
AMgvattdhi strongly suggests a survival of the subjunctive seen in Vedic
vdrtds{i) (cf. the Skt. “ imperatives ” first singular vdrtdni, first plural
vdrtdma, which are really subjunctives). In the imperative forms with
-e- beside those in -a vattesu, vattehi, vatthnha, vattedha, vattehu —
one clearly has contaminations of the imperative with the optative
(cf. Skt. vdrtes, etc.). The Pali second singular imperative vattassu
can scarcely have arisen directly from *vdrtasva (cf. Skt. vdrtasva),
despite Geiger (p. 109), for this should give *vattassa. It evidently
represents a contamination of vdrtasva and vdrtatu.
The first plural type vattemha (M, S, Mg, Dhakkl, and frequently
JM) is well explained by Pischel (pp. 332-3) as based on the
“ injunctive ” of the -s-aorist seen in Vedic gesma (i.e. ge-s-ma), jepna,
and desma, while the type valtaynha would, correspondingly, be for
*vdrtd-s-ma, the first plural of the Pali aorist avattamha beside avattdmo
(cf. Skt. dvartdrna) possibly pointing to an identical origin (for another
explanation see below, p. 572). The Pali second plural middle vattavho
is clearly for *vdrtadh{u)vas (Bloch, p. 232), a form which seems best
explicable as a contamination of the second plural *vdrtadhva (cf.
the unique Vedic ydjadhva, RV vni, ii, 37, and Greek (fyipeaOe [see
Brugmann, ii, iii, 651, and Gray in Language, vi (1930), 239]) with
OBSERVATIONS ON MIDDLE INDIAN MORPHOLOGY
571
a second singular active “ injunctive ” corresponding to Skt.
{d)rartas. The third plural vcUtahl, like the second plural vattaha,
is merely the indicative used as an imperativ'e.
In the future the normal ending of the first singular is -m rather
than -mi (Prakrit vattissam, Pali valtissam, as contrasted with Skt.
vartisydmi (Gatha Pali even has the type vattam in the present).
Here we seem to have a trace of the “ injunctive ” (cf. Pali aorist
agamisam and the Skt. conditional dvartisyam, Pali avattissam, and
see Bloch, MSLP, xxiii, 119), a similar phenomenon recurring in the
(rare) M verse-form of the first plural vattissdma beside the normal
vattissdmo and in the regular Pali type vattissdma (contrast Slct.
vartisydmas), possibly in the Prakrit first plural vattikissd, and probably
in the second plural vattihittkd.
Por the first plural, Prakrit grammarians record the types vatiikissd,
vattehissd beside vattissdmo, vattissdma, vattissdma, vattihdmo, vattihdmu,
vattihdma, vattihimo, vattihimu, vattihimvi. The termination -ssd
is obscure, but phonologically it may come from -tsd, -tsd, -hsd, -hsd,
-hsd, -syd, -srd, sld, -svd, syd, -svd, -syd, -svd (Pischel, pp. 217-18,
224-5, 225-6). Of these, only -svd {-svd) appears to be even remotely
possible here, so that vattihissd would seem to be for *vartisyisvd.
The form seems to be, in reality, dual, not plural, in origin, and appears
to be a contamination of the true future seen in Skt. vartisydvas,
vartisydmas, Prakrit vattissdmo, etc., with the -s-aorist (cf. Skt.
d-sraut-s-va and the M, etc., first plural imperative vattemha ~ Vedic
gesma cited above), with the long vowel which appears in the
Lithuanian reflexive tj’pe dirbavo-s. With the grade in e instead of 6,
we may perhaps also cite Old Church Slavic veze-ve in this connection.
This form is usually explained as influenced by the second personal
pronoun dual ve (see, e.g., Meillet-Yaillant, Le Slave commun^, Paris,
1934, p. 325), but it is at least possible that the analogy worked in the
reverse direction, especially as only Slavic shows a long vowel in this
pronominal form (cf. Brugmann, ii, ii, 412,426-7). The -h- in vattihissa,
etc., wonld seem to represent “ a special treatment of -ss- > -s- in a
terminational element ”, seen also in the Mg genitive singular pwfiaAa,
A puttaha, puttaho < *puttdsa < pultassa < putrdsya, *putrdsas
(Turner, JRAS 1927, pp. 232-5 ; BSOS V [1930], 50 ; VI [1932], 531).
The second plural of the Prakrit type vattihitthd (known only from
grammarians, with the exception of AMg ddhittha as contrasted with
Skt. ddsydtha) is interpreted (e.g. Pischel, p. 363) as based on an
“ injunctive ” of the -fs-aorist (Prakrit vattihitthd < *vdrtisyista).
572
L. H. GRAY —
We should, however, expect -tth-, not -tth-, < -ist-, just as we should
in the Pah second plural aorist active avattitiha, second singular aorist
middle avattittho, and third singular avattittha, Prakrit avaliitihd,
Asokan vadhithd. The presence in Pah of the tj'pes of avattattha
beside avattittha, with -tth- < -st-, sugggests that in the forms with
-itth- either we have a contamination of the -is-aorist with the
“ thematic ” aorist (i.e. the type of Skt. dbodhista contaminated with
that of dsicata ; cf. K. F. Johansson, in KZ xxxii [1893], 450-3,
though he considered the contamination to be with the -s-aorist
rather than with the “ thematic ”) ; or that, as suggested by Mr. Smith
(cf. also Henry, p. 89 ; Bloch, p. 232), the termination has been
influenced by the second plural of the copula (M ttha, Pali attha, Skt.
sthd), which, like as-, sthd-, vart-, and hhav-, had already been used
in Sanskrit with the present participle to express continuous action
just like English I am standing. Late Latin stat spargendo “ he is
scattering ”, Spanish estoy comiendo, Italian sto mangiando “ I am
eating” (cf. J. S. Speijer, Sanskrit Syntax, Leyden, 1886, pp. 294-5).
It is quite possible that both factors were operative simultaneously.
If this explanation may be accepted for the type vaUihittha, vattihitthd
would seem to have been similarly influenced by the vanished aorist
of sthd-, *atthd, corresponding to Pali atthd = Skt. dslhdt, the third
singular being generalized just as in the case of Prakrit dsi = Skt.
dsit (cf. Pischel, p. 359). Beside vadhithd Asoka has the normal form
nikhaniithd. For the cerebral in Pali aUhd (cf. M, JM thdi) see Geiger,
P-71. '
When we turn to the aorist, only Pali forms need renewed con-
sideration here. The first plural active of the type of avattaniha
< *dvdrtasma may derive its ending from the copula amha (cf. M
mho, mha, Skt. smds ; see Bloch, p. 232), though it is more probably
connected with the t}pe of such Vedic aorists as gesma, jemna, and
depna (see above, p. 570). The rare type of avattimha is the precise
equivalent of Skt. dvartisma, and the second plural avattittha is for
*dvartasta. For the second plural beside < *dvartasta
and dvartista respectively, and for -tth- instead of -tth-, we have already
advanced two hjpotheses in discussing the future (p. 572) ; and in
the t\pe of avattatha we have simply a borrowing of the present ending
{vattatha ; cf. Bloch, p. 232).
In the third plural types of Pali avattimsu (Prakrit avattiriisu),
-inisu is explained as equivalent to Skt. dvartisur (Pischel, p. 360 ;
cf. Henry, p. 89), but it seems, rather, to be a contamination of
OBSEEVATIONS ON MIDDLE INDIAN MORPHOLOGY
573
*avattum and *avattisuni (cf. Skt. dvrtan : dvarti^ur), i.e. of the -a-
aorist and the -is-aorist. Asoka has the forms alocayisu, nilchamisu,
ichisu ; cf. Skt. dbhavisur, etc.
The Pali second singular middle type vattittho cannot be compared
directly with that of Skt. dvartisthas despite Geiger, p. 130, but seems
to be from *dvarlisthds through confusion of the middle termination
-thas with the active -as (cf. Bloch, p. 231, and for the dental instead
of the cerebral, above, p. 572), together with a contamination of the
-a-aorist and the -fs-aorist (cf. Skt. dvartas : dvartisthas). The third
singular of the t\"pe of avattatha (Asokan Prakrit nikhamithd, vadhithd),
with -th- where one would expect -t- (cf. Skt. dvartata), is very puzzling
(cf. Bloch, p. 232). The only suggestion that I can olfer is that we have
here a contamination of the “ thematic ” with the -w-aorist, i.e.
avattatha < dvartata *avattattha < *avattasta ~ *avattittha
< dvartista. The quantity of the final vowel — short in Pali avattatha,
but long in Asokan vadhithd — may be due to analogy with similar
phenomena already observed in the future. The same -th- reappears
in the Pali third singular optative and conditional middle {vattetha,
avattissatha), and may conceivably be due to the fact that these, like
the aorist itself, are characterized by “ secondary” endings.
The termination of the first person plural of the t\'pe of avattamhase
can scarcely be derived directly from *-mase, as stated by Geiger
(pp. 107, 109). It must stand for *dvartasmase, which seems best
explained as a mere borrowing from the present type of vattamhe
(see above, p. 568).
The A infinitive of the type of valtana is ob\dously the nominative
singular of an action-noun corresponding exactly to Skt. vdrtanam
(for the use of the bare stem instead of the form in -m cf. Pischel,
p. 248), and it surxdves in Modern Indian (Bloch, p. 283). It can,
however, scarcely be equated with Kanesian (“ Hittite ”) infinitives
of the t}'pe of hark-ana “ go to ruin ”, since these are probably datives
beside the regular t}pe of idalawani “ to injure ” (E. H. Sturtevant,
Comparative Grammar of the Hittite Language. Philadelphia, 1933,
pp. 151, 267). Beside rattana we also find vattanahi (probably a locative
singular, less plausibly a locative or instrumental plural ; cf. Pischel.
p. 392) and vattanahd, which looks like a genitive plural (Pischel.
loc. cit.). Since, however, no infinitive in the plural is known in Indo-
European, it may be suggested that vutlanahd is for *vdrtanatham-
i.e. a noun of action in -no- plus the Indo-Iranian formative -tha-
< -tho-, also used to make nomina actionis (e.s. Skt. gdthd-. Av gd6a- ;
574
L. H. GRAY—
Skt. hhrthi- ; cf. W. D. Whitney, Sanskrit Grammar^, Boston, IJ.S.A.,
1896, p. 436 ; B. Lindner, Altindische Nominalbildung, Jena, 1878,
pp. 84-5 ; Bartholomae, pp. 107, 110). At the same time, -tha- is not
used either in Sanskrit or in Avesta to form secondary noim-derivatives,
so that the suggestion is open to grave objection. Another
A infinitive of the type of devam “ to give ” would seem
to be the locative (cf. Vedic mfirdhdn beside murdhdni, and see
Brugmann, n, ii, 177-8) of the formation in -yen- which appears in
inflected form in infinitives of the type of Vedic ddvdne, Cyprian
hofevaL, Greek Sovyai, GAv mdvanoi, and which finds an exact
counterpart in Av roi6w3n and the Kanesian “ supine ” type of
peskewan “ giving ” (Bartholomae, pp. 145, 146 ; Sturtevant,
p. 153).
The Pali infinitival types vattildye and vattetuye, like the AMg
vattae, vattitae, are obviously datives (cf. Bloch, p. 252 ; cf. also the
survival of the dative in Pali in an infinitival use, e.g. dassanaya
“ to see ”, Geiger, pp. 78, 151). On the other hand, vattae, etc., are
scarcely to be equated, despite Pischel, p. 392, with Skt. vdrttave,
even if intervocalic v occasionally vanishes in Prakrit (Pischel, p. 137),
but must be exactly equivalent to Skt. vdrtaye. Similarly, the Pali
type of vattitdye would seem to be a contamination of *vdrtitaya
and *vdrtitaye (datives of *vdrtita- and *vdrtiti- respectively), just as
vattetuye is of vdrtayitmn and *vdrtayataye (i.e. the causative infinitive
and the dative of *vdrtayati- ; cf. Bloch, p. 132).
Of the gerunds (or absolutives), the types of JS vattiduna, Pallava
Grant vattitunam, JM vattiuna(m), Gatha Pali (rare) vattituna,
represent a contamination of a gerund in *-lvana (cf. Vedic -tvdnam,
cited only by grammarians) with the infinitive in -turn, with -tu < -tu
by analogy with *-tvdna and probably also with gerunds of the type of
AMg vattiydnam, Pali vatliydna (cf. the Vedic alternation here of -ya
and -yd, e.g. ahhikrdmya : sarhgrbkyd ; A. A. Macdonell, Vedic
Graminar, Strasbourg, 1910, p. 413 ; see also Pischel, p. 400 ; Geiger,
p. 156 ; Bloch, p. 284) ; i.e. *-tuna{m) : *-tu < -tu = -ydna{m) :
-yd = -tvdnam : -tvd. The formation in -tvdnam actually appears in
the AMg vattittdnam, and beside it are vatticcdna(rh) and vatticcd,
with -cc- < -ty-. To assume a mere substitution of y for v in the two
latter, with Pischel (p. 397 ; cf. p. 177), seems very improbable. These
two types are, rather, to be compared directly with the Vedic gerund
in -tyd (e.g. elyd, abhijitya ; cf. Macdonell, p. 414), and they would
seem to be from *vajtiydna(m) < *vftityd, a late formation froni
OBSERVATIONS ON MIDDLE INDIAN MORPHOLOGY
575
*vft-tyd (cf. 8kt. vftya) plus a formative -{a)na{m), and from the simple
*yflitya respectivelv. The type of AMg vattiyanam. Pah vattiyana,
appears to come, similarly, from *vartiydnam < *vrtiyd, an extension
of a form precisely identical with Skt. vftyd (see Geiger, pp. 155-6)
plus the same formative -(d)n<i(m).
Much more problematical is the A type vaUi, but it is at least clear
that Pischel’s explanation (p. 401) — ^that it has arisen from the type
of vattia through loss of a final -a — is quite improbable. The question
is rendered still more difficult since Modern Indian shows that the
form must have been vatti (Bloch, p. 285). The only suggestion I can
now make is that we here have a survival of a locative infinitive in
-i, of the same type as the Vedic infinitive nesdni (only eight instances
recorded ; see Macdonell, p. 412) and GAv frarsni (Bartholomae,
p. 146). A like confusion between the infinitive and the gerund is
seen in the type of AAIg kdum — Skt. kdrtum as a gerund (cf. Pischel,
pp. 390-1, 392-3, 398). It may also be significant, as regards the
final -i, that one finds in Vedic (though very rarely) kartdri beside
kartdri, and vedl beside vedi (^Vhitney, pp. 120, 134, 138 ; Macdonell,
pp. 245, 271, 284). For other locative infinitives cf. Av. -tacahi,
Greek Doric < *e;^€ 0 ’€v, Homeric tSfiev, Latin
ayere < * age si, Av. raiOwan (Brugmann, li, iii, 899).
Finally, the Pali future passive participle of the type of vattitdya-,
vattitayya-, vattiteyya-, to be compared with Asokan pujetaya-, is
clearly a contamination of the types of Skt. vartilavyd- (in Vedic
only janitavyd- and kinisitavyd-) and Vedic vrtdyia- (e.g. viddyia - ;
cf. Macdonell, pp. 406-7).
The results obtained in the foregoing pages may be tabulated
as follows : —
Sg. nom. m.
Sg. gen. m.
Sg. abl. m.
PI. nom. m.
„ f-
I. XoiNS
Mg, AMg
A .
M, etc. .
putte, A puttu *putrd:, *putrbi
puttaho, *putrds{y)as
puttassu
puttdo *putrdtas
M . putidhi{mto)
A . . puttahe,
puttaku
. Alklg . puttdo
, M, etc. . devlOj vahuo
* M, etc. . phaldi{m),
A phalai
*pvirdbhi{mtas)
*putrdsyds,
*putrdbkas
*putrda8
*devias,
*vadhiias
*pkdldin{i)
putrdsya ~ *pntrds -f -as
pnfrdt putrdtas (or puirdt
-f -as ?)
putrdt ^ *pvtrdbhi{m){tas)
*putrd -f -as of cons, stems.
*vad}iu + -as of cons,
stems.
phdld + -in{i) of -in- stems.
576
L. H. GRAY —
II. Verbs and Verbals
Act. pres. ind. sing. 1
A .
vattau
*vdrtakam
vdrtami ^ ahakdm
„ 2
A .
vattahi
*vdrtadhi (?),
„ ,, plur. 1
A .
vattahu
vdrtasi (?)
*vdrtatham
vdrtaiha ~ A sing, vattau
2
A .
vattaku
*i'drtathas
vdrtatha ~ vdrtdmas
„ 3
A .
vattahi
*vdrtathim
vdrtatha ~ A sing, vattai
Mid. „ „ 1
Pali
vattamhe
*rdrtasme
vdrtdmahe ~ *asme ~ *asmas
Pali
. vattdmase
*Ldrtdmase
< smds
vdrtdmasi ~ vdrtdmahe.
Pah
. vattamhase
*vdrtasmase
Pali vattamhe ~ vattdmase
Act. pres. opt. sing. 1
Pali
vatte
*rdrtem
(for Type B see pp. .
Act. pres. opt. piur . 1
368-9)
Pali
vattemu
*idrtemu
vdrtema ~ vdrtatu
.. 3
Pkt.
vatte
*vdrten
As.
vattdvu
*vdrtdvur
vdrteyur • — < vdrtdn
Act. impv. sing. 1
As.
Pkt.
vattevu
[iattdmu\
*idrtevur
*vdrtdmu
vdrtdmi vdrtatu
Pkt.
[vattamu]
*vdriamu
(d)vartam • — ■ vdrtatu
.. .. 2 .
Pkt.
vatteisu
*rdrtasu
(d)vartas ~ vdrtatu
Pkt.
vattesu
*vdrtesu
vdrtes vdrtatu
A .
vaitu
*vdrtas
(d)vartas
A .
vattahi
*vdrtadhi
vdrta ~ type of sftiu{d)ht
Pkt.
vattehi
*vdrtedhi
*vdrte (!) > — ' type of sfnu{d)M
AMg.
vat^hi
*vdrtadhi
vdrtds(i) ~ type of sp),u[d)h\
Pali
vattassu
*vdrtassu
vdrtasva ~ vdrtatu
Act. impv. plur. 1
M, etc.
vattamha
*idrldsma
vdrtdma type of Ved.jeVjnu
M, etc.
vatUmha
*idrtesma
vdrtema ~ type of Ved. jesma
A .
vaUahu
*vdrlalhu
vdrtatha ~ vdrtatu
A .
vattehu
*rdrtetkn
*vdrtetha ~ vdrteta ~ vdrtatu
Mid. impv. plur. 2
Pali
vattavho
* rdrtadkras
vdrtadhva ~ (d)vartas
Act. fut. sing. 1
Pkt.
vattissarhy Pali
*iarlisyam
vartisydmi ~ (d)vartam or
plur. 1
Pkt.
vattissam
vattissdmat
* vartisydma
(d)vartisyam
vartisydmas ~ (d)vartisma
Pkt.
Pali vattissdma
vattihisdd
* vartisyisid
vartisydvas ~ (d)vdrtisva (cf.
» .,2
Pkt. .
vattihittha
*vartisyista
Lith. dirbavo-s, OCSl. veze-
ve)
vartisydtha ~ *(d)vartista ~
{d)vartata or vartisydtha
sthd
Act. aor. plur. 1
Pkt.
. vattikittkd
*vartisyistd
vartisydtha {d)sthdt
. Pali
. avattamka
*dvartdsma
dvartdma ^ type of Ved.
jesma or dvartdma *S7nd :
smds
Pali
. avattimha
*dvartisma
dvartisma
»7 ,, 2
Pali
. avattattha
*dvartasta
dvartata dvartista
Pah
avattittha
*dvartiMa
dvartista dvartata
»' »» 3
Pali
. avattirhsu
^dvartum
dvrtan ^ dvartisur
Mid. aor. sing. 2
*dmrtisum
. Pali
avattittho
*dvartisthas
dvartisthds dvartas
,, „ 3
. Pali
arattatha
*dvartntha
dvartata dvartista
As.
. avattitha
*dvartUha
dvartista dvartata
{d)sthdt
OBSERVATIONS ON MIDDLE INDIAN MORPHOLOGY
577
Mid. aor. plur. 1
Pali
Infin.
. A .
A .
A .
AMg
Pali
Pali
Gerund .
. Jg
AMg
AMg
A .
Part. fut. pass.
. Pali
avattamhase
*dvartasma^e
vat^na
*vdrtanam
vattanahl
^vdrtanasmin
vattanahd
*vdrtanatham
vattae
* vdrtaye
vattitdye
^vdrtitaye
vattetuye
*vdrtayatuye
vattiduna
*vdrtituna
vatticcdna{m)
^vartitydna{m)
vaftiydnaiht
Pali vattiydna
*vartiydna{m)
vatti
*vdrti
vattitdya-, etc. y
As. vattiUya-
*vdrtitdya-
dvartamahe ^ *dsme
*dsmas < smds
vdrtanam
(cf. loc. puttahi)
vdrtaye
*vdrtit':iya — •vdrtitaye
vdrlayitiim ~ * virtayataye.
•vrttvdna ~ vdrtitum
*vkityd < *v'rttyd +
-(d)M(m)
*v'rtiyd < v'rty& + -(a)na(m)
{cf. nesdni)
vartitavyd vrtdyia-
Sakische Etymologien
By Olaf Hansen
1. Sak. ggathaa
O AK. ggathaa „ Laie “ und seine Ableitungen lassen sich als
^ buddhistiscber Terminus kaum von skt. grhasthd trennen.
Offensichtlich deutet das sakische Wort auf ein mittelind. *gdttha
Oder *gahaUha (vgl. pall gakattha), aus dem es entlehnt ist. Derselbe
Laut th fur skt. sth findet sich auch im sak. Lehnwort vathayaa <
mitteUnd. uvatthdyaa, skt. upastkd-. Die Herleitung von sak. ggathaa
aus dem Mittelindischen scheint mir gegeniiber der von Konow in
seinen Saka Studies p. 136 gegebenen besonders aus dem Grunde den
Vorzug zu verdienen, als auch eine andere iranische Sprache das
Wort derselben Quelle entnommen hat : soghd. k'rtk Dhyana-Text
6 und k’rt’k Dhuta-Text 206, dessen prazise Bedeutung ich mit Herrn S.
Matsunami im Jahre 1932 aus dem chinesischen Paralleltext
erschhessen konnte. Wie Benveniste unabhangig davon nachgewiesen
hat {JAs. 1933 (2), p. 214), steht k'rtk, k'rVk im Gegensatz zu soghd.
kty’ky ^ nyztk „ pravrajita “ und zu hynh'r „ religieux “. In beiden
angefiihrten soghdischen Schreibungen ist rt offenbar ein Versuch.
mittelind. tth wiederzugeben. Es liegt hier demnach eine ahnliche.
Schreibung vor wie in soghd. pivrny^nyh „ punya “ (Rosenberg,
Bull, de I’Acad. des Sciences, 1927, p. 1391 f.), wo ind. n dutch soghd,
rn wiedergegeben wird. Daneben steht jedoch gewohnlich piimy'ny.
AhnUch wie im Soghdischen sind vereinzelte Schreibungen Alberunis
bei der Wiedergabe indischer Worte zu beurteden (vgl. Sachau,
Indo-arabische Studien zur Aussprache und Geschichte des Indischen
in der 1. Halfte des XI. Jh., Abhandlungen der Kgl. Preuss. Akad.
Wiss., 1888, p. 18). Alberuni schreibt in arabischen Zeichen rn resp,
nr fiir ind. n ^ in = banij, = Panioi (zu diesem Wort
vgl. Sachau l.c.), und rt fiir th resp. tth in ^ — kunatha (Sachau,
p. 17) und Cj jf- = skt. jyaistha (Uha < skt. stha).^ Von dieser Art
Schreibungen ist ganz zu trennen soghdisch yivrnw VJ. 783, das
' Betreffs kty'ky sci auf das allerdings noch nicht ganz sicher gedeutete soghd.
'gfnphy (yr'mth'rt) der Kara-Balgasun Inschrift Zeile 14 {vgl. Mem. Soc. Finno-
Ougrienne 44. 3, p. IS u. 32) aufmerksam gemacht.
Daneben gibt es jedoch Falle, wo Alberuni fur ind. y einfach n Oder auch r
schreibt (vgl. Sachau l.c. 18).
® Zj ist bei Sachau versehentlich unter die Beispiele gestellt, wo th mit o
wiedergegeljen wird, wofur Sachau mehrere Beispiele anfiihrt.
580
O. HAKSEN
Gauthiot richtig mit „ ensanglantee “ iibersetzt hat : rty ZKh pnv
’stJcwh ’vsprynt ’skwn KFH ’Pny k’lv ’nspr’ynt rty ZKh z’yh s't yitrnw
^’y d.h. „ und sie wanderten auf Knochen, sodass dort, 'wo sie
wanderten, die Erde ganz blutig wurde ywrnw ist Adverb auf -w
(*-u < *-am) vom Adjektiv ^yicrn- d.i. *xuriiu'in{a) oder ■x^irUyi{a)
mit der auch bei r belegten Umstellung des w : vgl. soghd. dwrt- = mp.
drut. Somit lasst sich soghd. ywrnw zu aw. yrrant- ,, grauenhaft
eigentlich „ blutig “ {yrura „ blutig, grausam “) stellen. Zur Ableitung
sei auf lit. kriivinas, abg. JcrbVbm- „ blutig “ verwiesen. In dem zweiten
Teil der Soghdischen Texte von FWKMiiller (hg. von W. Lentz) ist
Xwrny im Ausdruck ywrny psyncn’ „ (o) Blutvergiesser ! “ belegt,
das mit dem oben aus dem VJ. belegten ywrn- identisch ist ; offenbar
ist yvirn- auch hier als Adjektiv aufzufassen.
2. Sak. bir
Sak. hir- ,, werfen, streuen “ lasst sich mit lit. beriu, berti streuen,
lett. behi, bert schiitten, streuen, bdrstit wiederholt streuen, bdrsli
allerlei Abfalle, die man den Hiihnern als Futter hinstreut, bdrstal'a
jmd., der alles ausstreut, bdrstes ausgestreutes Getreide, bifstu, biru,
birt sich verstreuen, ausfallen, abfallen (s. Miihlenbach-Endzelin,
Lettisch-deutsches Worterbuch s.vv.), vereinigen. Die baltischen
AVorte verbinden Buga, Aistiskai Studijai 1908, 184, Mikkola, Urslav.
Grammatik 43, Briickner in : Gesch. der idg. Sprachwiss. ii, 3, 83,
und Trautmann, Balto-slav. AA'orterbuch 31 mit idg. *bher- ,, tragen
ohne eine nahere Begriindung fiir den Bedeutungswandel zu geben.
Einen A^ersuch, diesen Unterschied in der Bedeutung zu uberbriicken.
unternehmen Muhlenbach-Endzelin, l.c. p. 292, der jedoch nicht sehr
iiberzeugt. AA^alde-Pokorny verhalten sich dieser A'^erbindung gegeniiber
ablehnend (s. Vergleichendes AA'orterbuch der idg. Sprachen ii, 165).
Der A’’okal des sak. bir erklart sich durch Umlaut : < *barya-.
Demnach stimmt sak. hlr ganz zu lit. beriii, lett. befit (-ie/-io Praes. :
vgl. auch Endzelin Lett. Gr. § 625). Dieser Umstand, dass zwei
Sprachgruppen uberein.stimmend ein -iej-io Praescns aufweisen, das
nirgends von *bher- tragen bezeugt ist (vgl. jetzt die Ausfiihrungen
iiber *bher- bei Ernout-Meillet, Diet. Et. langue lat., pp. 334 f.), spricht
neben der Schwierigkeit, die Bedeutungen zu verkniipfen, fiir den
Ansatz zweier getrennter AA’urzeln. Uber weitere Ankniipfungen dieser
AA'urzel *bher- in anderen idg. Sprachen s. Aleillet, A propos de la
groupe litouanienne de beriii, Streitberg Festgabe, Leipzig, 1924,
258 ff.
SAKISCHE ETYMOLOGIEN
581
3. Sak. haema
Im sak. Bhadrakalpikasutra (hg. v'on Konow, Avhandlinger utgitt
av det Norske Videnskaps-Akademi, ii, hist.-filos. kl. 1929, No. 1,
p. 11) steht : (33) . . . u Jiarma beti ttaramdara vamnlha ni byehe, was
Konow I.C., p. 13 ,, and no poison finds entrance into the body “
iibersetzt. Im Glossar zu dem zitierten Text bemerkt er (l.c. p. 17) :
harnm unidentified perhaps „ any “. In seinen Saka Studies, 1932.
p. 141, gibt er mit dieseni selben einzigen Beleg : liarma of any kind,
any, nom. -»m, und vergleicht aw. Jiaurra. Gegen den Vergleich mit
dem awest. Wort scheint die Verschiedenheit der Suffixe zu sprechen,
besonders da in den iibrigen iranischen Sprachen dieses Wort in der
sakischen Gestalt bisher nicht belegbar ist. Die neuen tocharischen
B-Texte, die S. Lem 1933 (Fragments de textes koutcheens publics et
traduits . . . Cahiers de la Soc. Asiat., 1''' serie, ii) veroffentlicht hat,
belegen ein toch.-B. solme „ complet das Meillet Bull. Soc. Ling.
35, Comptes Rendus, p. 51, richtig mit toch. A salu, skt. sarvah usw.
vergleicht. Toch.-B sohm und sak. harma weisen beide das mit m-
anlautende Suffix auf und stehen damit im Gegensatz zu alien iibrigen
idg. Sprachen. Ob da zwischen solme und harma ein direkter
Zusammenhang besteht, der bei den verschiedenen gegenseitigen
Beeinflussungen dieser beiden Sprachen nicht ausgeschlossen ist, lasst
sich vorlaufig nicht entscheiden.
Soghdische Miszellen
Von W. Henning
I. Pynpt-
I N einem der von mir unlangst veroffentlichten manichaischen
Texte in mittelpersischer Sprache kommt zweimal ^ ein Wort
hsnbyd vor, fiir das sich die Bedeutung ,, Gotzenpriester “ zwingend
aus dem Kontext ergab ; die Herkunft des Wortes kannte ich damals
noch nicht : sie wird mit einem Schlage klar, wenn man soghdiscb
Pynpt- ,, Zauberer “ herbeizieht {pynptw SCE. 255 ; AUe Brief e
i, 10 ; fern, ^ynpt’nch BCE. 255, 252), das offensicbtlich mit mp.
hsnbyd etymologisch identisch ist.^
Die gemeinsame Grundform muss *baginapati - ,, Herr des hagina- “
gewesen sein, woraus in persischem Munde, mit Palatalisierung des
g vor hellem Vokal, *bazinapati- wurde. Nach Schwund des zveiten
Vokals ergaben sich die Formen baynapal (daraus soghd. ^ynpi- =
^aynpat-) imd baznapet, aus dem nach der im Persischen gelaufigen
Verwandlung® von -zn- in -hi- schliesslich basnbeB (=mpT. bsnbyd'j
werden musste.
Dass in verschiedenen iranischen Sprachen der alte Velar eines
und desselben Wortes einmal bewahrt, ein andermal hingegen
palatalisiert ist, kommt garnicht so selten vor ; es sei gestattet,
hierfiir ein paar Beispiele anzufiihren, die sich jedoch leicht vermehren
bessen :
(1) Parthischem ivrwc = viroc > viroz * Blitz “ steht im
Persischen lorwg ® = irirdy < *i'irok- gegeniiber (hierzu auch der
Eigenname mp. ivrwgdfd = Virdyddb ,, vom Bhtz gegeben “ in Manis
Gigantenbuch, = ,, YrnMad“ bei Bang Manich. Erz. 13 ss. ®) ; zur
^ Andreas- Henning, Mittelir. Manich.., ii («S6.P.A.U'. 1933), 311, 21 ; 312, 16.
- Die Form Pynpt- bei Benveniste Gramm., ii, 78, 89, 213, existiert nicht ; die
dort, p. 78, gegebene Erklarung ^aydn°pat- ,, maitre des dieux, sorcier “ ist daher
vollig haltlos.
® Cf. Meillet-Benveniste, Grarnm.^, 69.
* Andreas-Henning, Mittelir. Manich., iii (S6.P.A.ir., 1934), 908.
® In unveroffentlichten manichaischen Texten.
® Museon, xliv; das dort von Bang behandelte turkische Fragment, auf das niich
seinerzeit Professor Schaeder hinwies, gehort in den Kreis der Henoch-Geschichten,
die sich an Manis Gigantenbuch und an sein Henochbuch anschliessen, vgl. Sb.P.A.W.
1934, 27 ss. ; Qonuy Burqan {lies : ist Henoch.
584
W. HENNING
Bildung vgl. ai. virokd-, viroJcin- usw. ; die im Ai. so gelaiifige
Komposition von v" rue mit vi ist im Iranischen fast imbekannt ;
ausser urwe usw. gibt es bier nur das denominative Verbum partb.
urusn’dn ,, aufleucbten “ i = sak. barun- ,, to sbine
(2) Partb. ’bg'iv- = afydv- ,, vermebren “ < *abigdvmja- gegen
ap. abizdvaya- (woraus sogbd. = a^zdv- und mp. 'bz'y-
= a^zdy-), die zur Wurzel av. gav^ ,, versebatfen “ geboren.®
(3) Mp. a^dy ,, mit “ aus *iipdk- gegen judpers. a^dz aus *updc-.*
(4) MpT. dvdy ,, Stimme ‘‘ ® (np. dvd „ Ruf, Sebrei “ Sn. 81, 367 ;
Asadi 57, 1, 2, und sonst) < *dvdk- gegen Pebb, np. dvdz < *dvdc-.
(5) Sogbd. mrc ,, Tod “ = marc gegen av. ma{h)rka-, np. mdrg.^
(6) Buddb. -sogbd. erks „Geier“ (Frag. Ill 28), oss. cdrgds
(Reicbelt I, 63, N. 3) gegen av. ka(k)rkdsa-, np. kdrgds.
Das erste Gbed des mit ziemUeber Sicberbeit rekonstruierten
altiranischen Kompositums *baginapati-, also *bagina-, bedeutete
,, Tempel wie sich aus dem zu arsacidischer Zeit vom iArmenischen
entlehnten Wort bagin ,, Tempel, Altar “ zweifelsfrei ergibt ; *bagi-
napati- ist also der ,, Tempelherr “. Was nun die Herkunft von
*bagiva- anlangt, so ist es wohl klar, dass dies Wort ein substantiviertes
-iba-Adjektiv zu baga- ,, Gott “ ist; wie av. hqniina- als Adj.
,, sommerlich “ und als Subst. ,, Sommer “ bedeutete, so hiess *bagina-
,, den Gottern zugehdrig “ und substantiviert ,, das den Gottern zuge-
borige, der Tempel ‘b Ausser dem Armeniseben sebeint nur das
Sogbdisebe dies M'ort bewabrt zu baben : es kommt, in der Form
Pyn-, mebrmals in einem nocb niebt veroffentlicbten maniebaiseb-
sogbdiseben Text der Berliner Sammlung vor ; bier eine
ebarakteristisebe Stelle (T II D 117 B V 11 ss.) :
’rty bywyh 8y^nw Pyiiy’ cyndr civ ^yysU ’skumd s't uysprlriynyt
xnd pyst mrtxmytyy pryw nyy j'ynd ’rty qfni- ”8yy ’ww ru'8nyt ^yyst
ky ty wy ^yny’ ’skwnd ’xswvdyy [Abbrucb] ,, Und wieviel Gotzen es
in diesen beiden Tempeln geben mag, die sind zwar alle mit jeglicber
Art Edelgesteins gescbmiickt : allein, mit den Menseben reden konnen
^ M. 81, R. 4, wrivsrid (unveroff.).
" Konow, Saka Studies, 12o.
^ S. Mittelir. Manich., iii, 892, s.v. 'hgivdn; ebenso jetzt Schaeder, Sb.P.A.W ■
1935. .502 X. 3.
s
«
{mrti
Vgl. zuletzt Xyberg Hilfsbuch, ii, 11 ss.
z.B. Mittelir. Manich. i {Sb P A.W 1932) 204.
Dies lieispiel ist freilich nicht ganz sicber, da sogbd. mrc auch auf av.
-) zuruokgehen konnte (mit ti > <5) ; anders Gauthiot, Gramm, i, 94.
SOGHDISCHE MISZELLEN
585
sie nicht. Ferner, wer auch immer die metallenen ^ Gotzen, die im
Tempel sind, zufrieden ^ [zu stellen sucht] . . .
11. zwrnyy
Buddh.-soglid. ,, nwrn’h ,, mvrn’y das bisiier meist falschlich.
mit ,, glaubig “ iibersetzt vnirde, bedeutet ,, Epocbe wie Lentz
soeben iiberzeugend nachweist.® Freilich geht die von Lentz
vorgescblagene Verkniipfung mit buddb.-sogbd. nwr ,, jetzt “ fehl :
das Wort ist vielmebr zicr7i- zu lesen und zu av. zrvan- Zeit auch
„ Zeitabschnitt zu stellen. Den Beweis dafiir liefert das Manichaisch-
Soghdische, wo das Wort in der Form zwrnyy nicht selten vorkommt ;
es sei gestattet, eine eindeutige Stelle anzufiihren (T II D 115 V 1 ss.,
unveroff.) : z'wr B^rnd skwn ^yystt wnyy mrtxmyty oo p’rt[y] Byn ’ty
rsty'kfrys’nd sk\w7i\ pr zwrnyy zwrnyy yr^’qy’ ywcnd skwn iryzryy r’BS
’mt’ynd skwn ,, Die Gotter geben den Menschen Kraft ; denn sie
senden, in den einzelnen Zeiten, die Religion und die ,,Gerechtigkeit
sie lehren die Weisheit und zeigen den wahren Weg “ ; bier entspricht
pr zwrnyy zwrnyy genau persischem pd w'm 'w’tn* In manich. Texten
begegnet auch eine Adjektiv-Ableitung von zwrnyy, namlich zM’mt/cyA:/
^ Vorher ist von goldnen und silbernen (nktynyy) Gotzen die Rede.
^ 'xiwndyy ist wobl dasselbe wie man.-soghd. 'xwsndyy ; of. M. 135, A. ii, R. 15 —
V, 2 (unverofif.) ; prw xict'wt 'xS'wnh'rtyh 'sp'sw’nw 'ly 'xwsiidyy skwynd ,, (Teile den
Tag in drei Teile, erstens) in den Dienst der Herren und Machthaber, auf dass sie
gnidig seien “ ; hierzu gehdrt wohl buddh. 'ytos'ni TV. 54c ,, froh “ (statt 'yics’nt
ist wohl ofters ’yws’nt zu lesen!) und sicher buddh. 'yws'w ,, ich hatte Freude
VJ. 1108 (Impf. sg. 1 wie das daneben stehende p’r’it’ ,, ich nahrte pt'yywsw TV.
430, 577 usw. ; von Gauthiot nicht erkannt).
MuUer-Lentz, Soghdische Texle, ii (56.P..4. TT’. 1934), 604. Ich benutze die
Gelegenheit, die Besprechung einiger in dieser Publikation vorkommender Worter
hier nachzutragen : ’wzb’ ,, Schrecken p, 578a ; dazu buddh. tvzp-yiim'kw TV. 313,
fern, wzpywnch T V. 63e ,, schrecklich “. yr yk- ,, Schlamm p. 583b (auch SCE.
160), gehort zu np. xdrd ,,Schlanira“ (daneben auch j A-, z-B. Asadi 25). qwrty
,, Gewand p. 586c ; hierzu gewiss russ. KvpXKa „ Joppe ", das wohl auf dera Uraweg
iiber eine Tiirksprache ins Russische gelangt ist und schliesslich auch im Nhd. als
,, Kurtka “ erscheint (z.B. bei Chamisso haufig). pst ,, Lippe ", p. 593a ist jungere
Form von buddh.-soghd. prst SCE. 86 (ebenso christl. p.sCd't von buddh., manich.
prst't usw.). su.'d't ,, eilen " p. 596b ; dazu buddh. su'S swSii TV. 1092 ,, eilends zu
eilen “ (auch manich. swS-) ; ob etwa hierher Phi. Ps. 124, 5, swd'tysn, das „ er wird
sie fiihren “ heissen soil, gehort? wy'wnt p. 600a: SCE 90 steht wy'wnty' (auf
dem Facsimile ganz deutlich), nicht wy wty . yx>«^l, P- 601b, gehort wohl zu yxw'y-
(christl. yu’x'y-) teilen, trennen “ wie ptx>xst zu ptyu-’y- usw., und nicht zu ywc-
,, lehren “ ; Pt. von ywc- ist buddh. yywt- < ywyt- „ gelehrt, gewohnt " (TV. 310,
320, 796 bzw. Dhuta 233; vgl. SCE. 536 yywsty „lernt“[so!] < ywysty Dhuta
162, 275).
* S6.P.A.IT'. 1934, 3.
* Auch buddh. : pystrw zwm’k cyk mrtym’k, Dhydna 369, ,, der spateren Zeit
angehorender Mensch “ ( =chin. homme fulur).
VOL. vm. PARTS 2 AND 3.
38
586
W. HENXIKG —
cf. T II D 207 R 1 s. (unveroff.) : III zwrnycylct pnc bictyst ,, die den
drei Zeiten angehorenden fiinf Buddhas ^ Neben dem vom obliquen
Kasus (mit Metathese) abgeleiteten zwrnyy besitzt das Soghdische
zwei auf den Nominativ av. zrva zuriickgehende Formen ^ : znv’ ,
"zriv’ als Gottesname (cf. mp. zrw'n dass.) und znvh „ Alter “ (cf.
parth. znv'n dass.).
III. Drei soghdische Worter im tiirlcischen Chvastvaneft
(rt) buhdunc
Der Satz rnunday buhdiowsuz^ yazuq yazintimiz drsdr ,, wenn wir
solche von der Erlosung ausschliessenden Shnden getan haben ‘‘
{Chv. I C Ende ed. Bang) wird in der verniutlichen soghdischen
Vorlage (entsprechend deni Stil nianich.-soghdischer Beichtspiegel
verwandle ich die 1. PI. in die 1. Sg.) ; ctv m'n’ piv bicxhonc yiv’n
'k-tu'8'rn gelautet haben, wobei *bu'rtwnc ,, Erlosbarkeit “ ein
c-Abstrakt zu *bwxtivnd ,, erlosbar, der Erlosung teilhaftig “ (<air.
*huxtivant-) ware. Abstrakta von raui-Adjektiven sind im nianich.-
soghd. gar nicht so selten, vgl. z.B. u'rcxicndqy’ ,, Wunderbarkeit “
(M 178 A R 18, unveroff. ; zu av. varocahvant-) oder das genau zu
*bu'Xtu'nc passende frnxicndc * ,, Glucklichkeit “ ; da nun das Abstrakta
bildende c-Sufids sich ohne Zwischenvokal unmittelbar an den letzten
Laut des zu erweiternden Wortes anschlie.sst, wie sicli z.B. aus bsi'wc ®
,, Armut “ < 8Mw'71 + c (mit Schwund des schhessenden Nasals)
ergibt, so verschmilzt leicht mit dem Suffix ein auslautender Dental,
der dann auch nicht mehr geschrieben zu werden braucht (man
vergleiche die femininen Partizipialformen manich. ’usuyc ,, rein “
51 178 A R 10 zu buddh. 'u'snytli, manich. „verniischt“ ibid.
B V 22 zu buddh. ptrystk, manich. ptsyc’ ,, geschmiickt “ ibid. A V 7
zu buddh. pts'ytk gegeniiber buddh. fern, pts'ytch YJ 16b, 40c, 1169,
1258) ; ahnlich wie bei *bwatu'nc liegt die Sache vielleicht bei christl.
’u'Xicnc ® ,, Kampf “.
^ Sachlich vgl. Chavannes-Pelliot, Traite^ ii {JA 1913, i), 137 N. 3.
^ Cf. Muller- Lentz, S.T. ii, 602b.
® So ist statt hu aduno'fiiz zu lesen, wie h bei einer von Professor Schaeder und
Frl. Dr. v. Gabain veranstalteten Universitatsiibung uber das Chvastvaneft lernte.
^ Waldsohmidt -Lentz Dogm, 548 V 22, und 580 ; daneben gibt es aiuehfrnxtvndkf/
(T II D 163 K 2; T II D 207 K 10), jrnxviiidtky'h (M 116 V 12; samtlich
unveroff.) : das legt die missliehe Vermutung nahe, dass frnxicndc aurh fern.
Adjektiv sein konnte.
^ Waldschmidt-Lentz, loc. cit. ; manich. auch bystjvc geschrieben, z.B. M 178
A V 25 ; ziim Vokal der ersten Sill>e vgl. christl. dlstvdn arm Miiiler, S.T. i, Hi,
16, pi. ibid. 34, 6, das Benveniste, Gramm, ii, 168, ubersehen hat.
® Muller-Lentz, S.T. ii, 577.
SOGHDISCHE MISZELLEN
587
(b) hoytay
Dass ,, biigtag “ in buyan biigtag qilurb'iz (Chv. viii B Bang) iind in
buyanc'i biigtagci ^ any di'ntarlarqa . . . yazintimiz drsar {Chv. iv, B)
aus dem Iranischen entleknt ist und zu nip. biixt usw. geliort, liat Bang
richtig erkannt.^ Freilich verbieten die Schreibung niit g und die
Bedeutung, an das passive Priiteritalpartizip mp. bnytay sellist zu
denken ; naber liegend erscbeint die Annalinie eines Kompositunis
mit Verbalnomen im Hinterglied : parth. boy ,,Erl6sung'‘ (buddli.-
sogbd. ^iiy Dhuta, 201, ^wlc gesclirieben, SCE. 215) + -tdy .. in
Bewegung setzend ‘‘ (causs. zu a,v. \/ tak ,, laufcn “) ; cin solches
*bdytdy ,, was die Erlosung bring! “ liesse sicb dann unniittelbar niit
parth. muMay-ddy .. Bringer des Evangeliunis “ (Waldscbmidt-Lentz
Stellung Jesu 86 ; ini Persischen entspricht mizday-tdz) vergleieben
und wild wobl urspriinglicli wie dieses dem Parthischen angehort
haben. Das von *b5ytdy, woraus tUrk. boytay (biigtag). abgeleitete
tiirk. nomen actoris boytayci (biigtagci) bedeutet demnach
,, Erlosungbringer
(c) bacaq
bacaq ,, Fasten " (im Chv. und auch sonst baufig). das im Turkischen
keine befriedigende Ableitung hat. diirfte wie so mancher Terminus
der turkischen dem Soghdischen entnommen sein, bacaq ist
wahrscheinlich eine tiirkische Weiterbildung von soghd. *pdc, das
aus air. *pdti-, dem nomen actionis von nv. \/ pd ,, hiiten, sich in
Acht nehmen entstanden ist ; der Lautiibergang von air. ti zu
soghd. c bietet keinerlei Schwierigkeit : er liegt vor z.B. in manich.-
soghd. j.Kauf" (xryc 'ty pr'Sn ,, Kauf und Yerkauf" M 135,
A II V 9 s., unveroff. ; dazu buddh. yr'yck' „ feil Reichelt i, 10,
153 ^), das auf air. *xrlti- (= skr. -krili-) zuriickgefuhrt werdeu muss,
und in dem haufigon Yerbalpriifix pc-, das gewiss nicht auf *pati.^
oder *pa6i-. wie gewohnlich angenommen ward, sondern einfach auf
pati beruht.^ Dass ein Derivat der Wurzel pd ,, hiiten “ die Bedeutung
,, fasten " haben kann, wird durch zalilreiche Beispiele bewiesen :
cf. nipT. pdhan mit und ohne ® rozny ,, fasten " (die Bedeutung
^ Diese Lesart clurfte vorzuziehen sein,
- Mnnichdische Laien-Beichtspiegpl {Museon, xxwi), 201 ss.
^ Danebeii auch yf ych dass. SCE. 356.
^ Die Bedin^imgen, unter denen pati zu pac wird, sind freihch noch jiicht klar.
^ In iinveroffentlichten Texten.
588
SOGHDISCHE MISZELLEN
„ Fasten “ von rozay stammt ja erst aus der Verbindung rozay paSan),
arm. pahk‘ ,, Fasten pahem ,, fasten “d
Sogbdisch *pdc ist jedoch dem Anschein nach von einem nacb
Gauthiot aus dem Indischen stammenden buddhistischen Wort
verdrangt worden : buddb.-soghd. ^ws'ntk (SCE. 77, 387), ^ivs'nty (ibid.
381). Dies Wort ist dann aucb in die Sprache der soghdischen (und
schliessbch der tiirkiscben ^) Manichaer iibergegangen ; es erscbeint
z.B. in einem Kolophon (M 214 E II 7, unveroif.) : ptyms ^wsndyy
nsk[rt\ ,, Es ist zn Ende : Die Deutimg des Fastens “ und aucb
(ganz wie im Chvastvaneft) in der Verbindung mit ,,yitnki“ ^ : ms
«[«/’ 3‘rS’ ymgyy ^icsndyy ’fr[y}ii:n p’syq ,, Seid ferner eingedenk
. . . des FawMiy-Fastens, der Gebete und Hymnen “ (M 134 I E 11-14,
unveroff.), vgl. CJiv. xv C, wo statt yimki bacaq die Petersburger
Handschrift yimki wusanti hat. Statt ytngyy, der soghdischen Form
von parth. yarmy, schreibt man auch ymqyy : pr myhr m'x jmmv
mrysysn ymqyy „ zur ,Mondzeit‘ (= Neumond) des (Monats)
Mihr findet (die Gedachtnisfeier) fiir den dpxyyos Mar Sisin statt “
(M 197 II V 3 s., unveroff.).
* 8. Hubsohmann Arm. Oramm. 217.
^ s. Bang, loc. cit., 218 s. ; neben dem uber das Soghdische entlebnten wusanii
gibt es im Uigurischea eine wohl direkt aus skr. upavasaihah stammende Form, die
ich durch Professor Rachmati kennen lernte, namJich 'posal {husat).
® t?ber yimki s. Schaeder [ranicay 22 ss., der Entlehnuiig aus einer jiingeren
parthischen Adjektivbildung *yamayl zu Unrecht annimmt ; das turk. Wort stammt
vielmehr wie fast alle iranischen Lehnworter in dieser Sprache aus dem Soghdischen,
namlich aus soghd. yfnqyyy ymgyy. Das von Schaeder herangezogene ymg nyg im
Petersburger Fragment S ist nicht eine adjektivische Ableitung vom Plural ymg n,
sondern, wie ich glaube, ein -d/jiy-Adjektiv von yaniay wie parth. bg'tiyg von bg u.
dgi. Der im zitierten Fragment auf die ytng'nyg rwc'n folgende Satz : 'ivs ^st ’c
pd yk mr wysp gw^g kann m.E. nicht einen Eigennamen V ispgosay enthalten (Schaeder,
loc. cit., 5 N. 4) ; der Satz muss doch etwa heissen : ,, Und es (sc. das Buch) enthait
in einer Hymne alle gosay “ ; gosay gehort vielleicht zu np. gosd ,, Winkel
Der Tod des Kambyses : hvamrsyus amryata
Von Ernst Herzfeld
Beh. § 11. Ap. passdva kambuziya h^vdmrs'yus amr'yatd.
Akk. arki kambuziia mi-tu-tu ra-man-ni-m mi-i-ti.
Elam, iak me.ni ^ kam.jni.ci.ia '‘al.pi t\ii.mayi\.e.'ma
'‘al.pi.'k.
"rxlESEE kurze, in drei sprachen erhaltene satz der Behistun-
inschrift des Dareios ist viel umstritten, von historischer und
pkilologischer seite. Heute kann man das ergebnis von W. Schulzes
bewundernswiirdiger abhandlung ,, Der Tod des Kambyses “ ^
allgemein angenommen nennen. H. H. Schaeder schreibt dazu ^
,, Die frage, so sollte man denken, ist damit erledigt — aber in
AMI. 5 best man “ und es folgt das citat eines satzes, in dem ich
meine bier begriindete, abweichende anschauung vorweggenommen
hatte. Jedes ergebnis hangt von seiner zeit ab, die zeit fliesst, und
keine frage ist fiir immer beantwortet.
W. Schulze hat gezeigt, dass viele indogermanische sprachen einen
ausdruck wie ,, er starb seines todes “ besitzen und damit den tod
als naturbches ende des menschbchen lebens, den dem menschen
bestimmten tod meinen. Zu diesen ausdriicken zahlt er das altpers.
hvdmriyus und sagt : ,, Die weithin verbreitete und nirgends verblasste
formel schbesst mord und selbstmord unbedingt aus, nicht aber
nach antiker auffassung die verhangnisvobe, doch unbeabsichtigte
selbstverwundimg, die gerade in Herodots darstellung sehr wirksam
als iiberraschende erfiillxmg eines orakelspruchs auftritt. Kambyses
selbst, der gemass der nachstbegenden deutung bisher gehofft hatte,
in seiner eigenen residenz iv toIs MrjhiKolai. ’Ay^ardvoiaL
TeXevrdv yrjpaios, also als jardmrtyuh, muss mm als erster
erkennen, dass sein schicksal sich anders und friiher als erwartet
erfiillen soil : ivravda, in der fremde, in dem sjTischen 'AyPardva
Kap^vaea earl Tmr paypivov reAein-ai'. — Auch so ist er, recht
verstanden, -npo yfyvcrecos zwar, hicht aber npo poipas- vom tod
erreicht. Herodots bericht, so ungeschichtbch er in der ausraalung
sein mag, bringt den gehalt des Ap. wortes in voller reinheit zum
ausdruck ; unmittelbar fassen kann man ihn am besten durch eine
antike umschreibung : /card polpav aTredavev, oder fato suo obiit. ‘
* SbPrAkdW. 1912. p. 699 ss und 1918, p. 331 s.
“ SbPrAkdW. 1935. xix, p. 496, anm. 1.
590
E. HERZEELD —
Strabo sagt I. 2. 1 : ov irpoKeiraL Trpog arravTag dvriXeyeiv, dAAd
Tovs pkv TToXXovs idv, eKeivovg Se Siairdvj ovg iv roZg nXeicTToig
KaTOjpdcoKOTag tapLiv.
Mebrere punkte in dieser darstellung Schulzes berechtigen zu fragen
und widerspruch : bewertet er Herodots erzahlung richtig ? Kann
sie den gehalt des Ap. wortes rein ausdriicken ? Gehort das Ap.
wort wirklich zu der allgemeinen wendung, und kann diese die
unbeabsichtigte selbstverwundung einschliessen ? Endlich, was
sagen die akkadischen und elamischen iibersetzer, die anders als
Herodot, als zeitgenossen das wirkliche ereignis wissen und den
lebendigen sinn des Ap. ausdrucks kennen mussten ?
Herodots erzahlung ist nicht bloss in der ausnialung ungeschicht-
lich ; sie ist die zusaramenarbeitung von mehreren, mindestens
drei quellen, deren wert er nicht kritisieren konnte.^ Die rolle
des orakels von Buto zeigt, dass Herodot als bewunderer
Aegyptens in der hauptsache der aegvptischen iiberlieferung folgt,
die alles im spiegel des priesterlichcn hasses sah. Ueber Buto, wie
auch iiber Syrien, hatte Hekataios ausfiihrlich geschrieben, den Herodot
zwar nicht citiert, aber in friiher unerkanntem umfang benutzte. Auch
die art, wie er seine aegyptische quelle verwendet, ist durch seine
polemik gegen Hekataios bestimmt. Die Prexaspes-erzahlung und
die letzten worte des Kambyses zeigen. dass Herodot daneben auch
rniindliche mitteilungen von Persern besass. Seine gefiihlseinstellung
zu diesen mindestens drei incongruenten stoffen war sehr verschieden.
So gestaltete er schliesslich ein sehr ,, wirksanies “, aber ganz
griechisches drama.
Der Apis war von Kainby'ses am schenkel verwundet und daran
nach einiger zeit gestorben. Da fiir die aegyq)tische legende Kambyses
tod die rache des Apis ist, mussten diese beiden ztige in die erzahlung
eingefiihrt worden, in welcher gestalt auch das geriicht vom wirk-
lichen ereignis nach x\.egyq)ten gelangte. Das ist keine geschichtliche
nachricht. — Der ort des unfalls ist bei Herodot ein Agbatana in
Syrien. In seinen nachrichten iiber Syrien benutzt er Hekataios. Erne
verlegung von Agbatana aus Medien nach Syrien kann man Hekataios
nicht zumuten. eher Herodot die interpretation eines anklingenden
namens bei Hekataios als Agbatana ; denn der aegy’ptischen quelle
kann das ,, in Svrien ” nicht angehoren, da es dem orakelsinn wider-
spricht. Dass es kein Agbatana in Syrien gab, wusste Nikolaos von
> vgl. AMI. V, 125 ss.
DER TOD DES KAMBYSES
591
Damaskos, der daher in dieser erzahlung das ,, s\Tische Agbatana “
in die ,, syrische bauptstadt “ umdeutend ,, Damaskos " sagt. Weder
ein syrisches Agbatana noch Damaskos waren der schauplatz der
tragoedie. Das orakel von Buto nennt mit Agbatana den waliren ort,
das wirkliche und einzige Hangniatana in Medien : ,, In Agbatana
wird Kambyses sterben.‘‘ Der konig legt den doppelsinn erfreut aiis :
in seinem herrschersitz, also in besitz seines konigtums werde er als
greis *zardmrsyus, wie dem menschen bestimmt, fato suo obire. Aber
gemeint ist, dass er in seinem herrschersitz alles verlierend „ an dem
ihm nicM bestimmten tage “ in seiner siinden bliite den tod linden
werde. Den tiefen doppelsinn des orakels hat Herodot vbllig verdorben,
und daraus ein falsches und triviales wortspiel mit zwei gleichnamigen
stadten gemacht. Es ware ein seltsamer zufall, wenn eine so
ungeschichtliche darstellung den gehalt des inschriftlichen wortes in
voller reinheit zum ausdruck brachte.
Trogus Pompeius folgt halb Herodot, halb Ktesias, bestatigt
also nichts. Die dritte und letzte quelle, Ktesias, will wie immer
Herodot widersprechen und ist daher auch negierend von ihm
abhangig. Von den aegyptischen einzelheiten lasst er nur den
rest bestehen, dass Kambyses an einer zufalhgen verwundung
stirbt, aber auf einer fahrt nach Babylon. Der alleinige grund
dafiir ist, dass Herodot nicht recht haben soli ; derm die
babylonische reise ist ganz unbegriindet und widerspricht der
einzig sicheren geschichtlichen thatsache, dass der tod bei der,
wegen der usurpation des Magiers iiberstiirzten riickkehr' aus
Aegypten erfolgte. Ktesias, der sonst in den iilteren teilen der Persika
bruchstiicke einer halb vergessenen, halb in sage tibergesansenen
uberlieferung widergiebt, wie er sie am hole horte. hat iiber Kambyses’
tod gar nichts gehbrt, sondern verschleiert nur bewusst seine abhiingig-
keit von Herodot. Er ist keine selbstandige quelle, wie Schaeder nach
den worten ,, sowohl nach Herodot wie nach Ktesias " zu glauben
scheint, und wir besitzen nur Herodot und die Dareios-inschrift.
An Herodots erzahlung ist der stoff meist tendenziose, aegjptisch-
priestorliche entstellung ; die ,, wirksam iiberraschende composition
ist sein griechisches drama. Geschichte ist nur die thatsache, dass
Kambyses in kritischster lage umsonst versucht, da Persis verloren
ist. das andre kernland seiner macht, das agbatanische Medien zu
erreichen, und in oder bei Agbatana stirbt ; er wird an das orakel
von Buto gedacht haben !
Bei der interpretation der inschrift-stelle muss man daher jede
592
E. HERZPELD
beziehung auf Herodot streng vermeiden. Aus der inschrift allein
hatte niemand geschlossen, dass Kambyses infolge einer zufaUig
selbstbeigebracbten verletzung, wie moderne geschichtschreiber gesagt
baben, an tetanus gestorben sei. Die inschrift bestatigt die
authenticitat von Herodots aegyptischer version nicht ; selbst nach
Schulze schliesst ihr ausdruck diese version nur nicht aus ; und selbst
in dieser einschrankung muss er ihn ,, recht verstehen namlich
77/30 (j)vaeojs, nicht vpo /xotpag.
Der kurze satz der inschrift ist danach die einzige geschichthche
nachricht iiber Kambyses’ tod. Das ist wert und zauber aller
epigraphischen studien ; es ist etwas andres, bei Herodot zu lesen,
was man sich himdert jahre nach dem ereignis in fremden landern
erzahlte, oder bei Dareios, der alles erlebte, was wirklich geschah.
Ich beginne die untersuchung des inschriftlichen satzes mit dem
am meisten vernachlassigten, dem elamischen text.
Im elamischen ist ’‘alpi das gewohndiche wort fiir ,, schlagen,
toten, caedere, occidere das in Behistun allein einige 30 male
vorkommt, in folgenden formen, die ich nach alten studien von 1915
mit paragraphenund zeilennummernder elamischen version anfiihre : —
a) ’'alpi, 1. sg. praet.
fur ap. azanam ,, ich schlug (ein heer) “ in § 18, i, 72 ; § 19, i, 77 ;
§ 31, ii, 53.
fiir ap. avazanam ,,ich totete (einen mann) “ in § 17, i, 65 ;
§ 20, i, 81.
b) ^alpiia, desgl.
fiir ap. avdkinam ,, ich totete (menschen) “ in § 13, i, 43 ; i,
45 ; mit sop ,, als ich getotet hatte “ in § 16, i, 57.
fiir ap. azanam ,, ich schlug (schlachten) “ in § 52, iii, 49.
c) ’‘alpis, 3. sg. praet.
fiir ap. am ,, mein heer schlug “ in § 25, ii, 18 ; § 26, ii, 27 ;
§ 27, ii, 31 ; § 28, ii, 35 ; § 29, ii, 43 ; § 30, ii, 47 ; § 38, ii, 83 ;
§ 41, iii, 11 ; § 42, iii, 16 ; § 46, iii, 26 ; § 47, iii, 33/34.
fiir ap. avdza ,, er totete (einen mann) “ in § 10, i, 24 ; i, 24/25
mit sop ,, als er ihn getotet hatte “.
fiir ap. avdzana 3. pi. praet. „ sie toteten ihn “ in § 23, ii, 8.
d) ^alpis, nicht optativisch fiir 3. sg. opt. ap. avdzaniyd in § 13, i, 39
u. 40.
‘‘ Ora'tnasta ^ un ir ’'alpnsne fiir ap. ahuramazddtaiy zantd hiyd in § 61,
,, A.M. soU dich schlagen “.
’‘alpisne .manlca fiir ap. imper. zadiy „ schlage, tote “ in § 38, ii, 82.
DEB TOD DES KAMBYSES
593
e) ’‘alpis .manka imper.
fiir ap. 2. sg. zadiy ,, schlage (das heer) “ in § 26, ii, 23 ; § 29,
ii, 39.
fiir ap. 2. pi. mtd „ schlagt (das heer) “ in § 25, ii, 15 ; § 33, ii,
62 ; § 50, iii, 41.
Pwiwana ir ’‘alpis kutta . . . -tnara fiir ap. vivahanam zatd (2. pi.)
utd ava kdram . . . ,, V. schlagt und das heer . . .“ in § 45,
iii, 22 s.
/) Diesen activen formen steht die passivische gegeniiber : ’‘alpika in
appo p^rt'ia ^alpika fiir ap. tya h^rb'ya avazata ,, dass BarSya
getotet war “ in § 10, iii, 41.
Dieselbe form steht in der nachricht iiber Karabyses’ tod. Die
formen mit -k, -ka sind 1. passiva, z.B. :
marrik = addriv ,, wurde gehalten “ ; tnarrika = dgrbitd ,, wiirde
ergriffen “ ; tintek = aniyatd ,, wurde gebracht “ ; tirikka = adahiv
„ wurde befohlen “ ; kuktak — ,, (mein gesetz) wurde befolgt “ ;
kusika „ wurde gebaut “ ; rappoka „ ^vurde gebunden “ ; pepraka
„ wurde gelesen “ ; tallik = nipistam ,, geschrieben ; hutlak = asti»
krtam ,, ist gemacht worden “.
Bei intransitiven, dazu bei verben der bewegung hat 2. das A:-sufl&x
stark mediale bedeutung, z.B. :
sinnik ,, profectus est “ ; parik = pardrasa ,, gelangte “, cf.
,, nanctus est “ ; puttukka ,, er floh “ ; halhk = \/yaiid-,
,, concitatus “ ; topaka ,, ausus est “.
An rtwriri zu denken liegt nahe, aber das elamische verb meint
iiberhaupt nicht sterben, sondern immer ,, toten
Die beschiidigimg des elara. textes betrifft den ausdruck fiir
,, eigen, hva- Die reste sind 1. ein kurzer wagerechter keil mit
kleinem winkel daran, 2. nichts, 3. ein e, 4. rm. Was das war, geht
sofort aus dem unmittelbar folgenden § 12 hervor ; tu-man-e-m%
,, eigen Das e ist sachliches possessivnm, das suffix -ma ist
locativisch imd instrumental. Also heisst es :
,, darauf Kambyses toten-eigen-sein-durch wurde-getbtet (od.
totete sich).“
Die elamische version spricht also den selbstmord uneingeschrankt aus.
Die akkadische version.
Das verbum ist als i, 1, ,, sterben •' ; das permansiv me-i-ti,
mi-i-ti heisst ,, er ist, war tot “ ; „ er starb “ wird nur fiir unsere
‘ Nicht wie Bork MAOG. vii, 3, p. 13, erganzt kor-pi-e „ seine hand".
594
E. HERZFELD
stelle angesetzt und nimmt ihr etwas von der absonderlichkeit des
ausdrucks, auf dessen erhaltung es gerade ankommt. — Der stamm
i, 2, zu dem der hier gebrauchte infinitiv mitutu gehort, hat eine etwas
modificierte bedeutung. DeUtzsch, der in nnserem falle ,, selbstmord “
ubersetzte, bemerkt mit recht, dass es nicht ,, durch sich selbst
sterben “ heisst. Aber es ist ganz deutlich, dass der stamm i, 2
immer, im unterschied von i, 1. das plotzliche, vorzeitige, auch
gewaltsame sterben ausdriickt ; oft ,, sie sollen sterben ,, sie
wollen nicht sterben “ u.a., immer im sinn ,, vor ihrer zeit Der
infinitiv i, 2 mitutu wird so gebraucht : ,, ich (Istar) werde den tod
(mitutu, nicht mutu) des Ahseri herbeifiihren der dann ermordet
wird. Ausschlag giebt die stelle „ Kuturnahhunte ina um Id semtisu
wruhis xmtfit " ,, K. starb plotzlich an dem ihm nicht bestimmten
tage Der begriff simtu nahert sich gr. fioipa, lat. faturn. Und doch
ist die vorstellung grundverschieden ; es ist das natiirliche schicksal,
und dem steht der begriff Id semtu fUr das, was die Griechen tragiscli
Oder i^almos nennen wiirden, gegenviber. Das ist der ,, Antike “
ganz fremd.
Yom selbstmord des konigs Ursa von Urartu heisst es in den
Annalen Sargons z. 139 : mit seinem eigenen schwert, wie ein schwein,
durchbohrte er sein herz/' Der ausdruck faUt, bei allem grauen, unter
die definition des witzes und ist als solcher beabsichtigt. Er ist dabei
voll der altorientalischen verurteilung des selbstmordes. In dem
bericht der S'"™* Campagne spricht Sargon spottend so davon : mums
Id tibe emid ramdnsu. Thureau-Dangin ubersetzt ,,une maladie
incurable il s'infligea a lui-meme namlich den tod, gegen den kein
kraut gewachsen ist. Dabei nimmt Thureau-Dangin meines lehrers
H. \\incklers alte erklarung ^ von sruldsu emedu ,, sterben “ an. Der
seltsame ausdruck ,, auf seinem berge stehen (?) “, halb mythisch,
halb euphemisti.sch, ist da ironisch verandert, indem mums Id tibe
fiir saddiu eintritt ; zugleich ist ihm durch den zusatz ramdnsu der
sinn ,, sich umbringen '' gegeben. Man konnte, die Ironie ausdriickend,
iibersetzen ,, an einer unheilbarcn krankheit starb er freiwillig
An .sich bedeutet also saddsii eumdu jedes sterben, ob natiirlich oder
freiwillig oder gewaltsam. Der begriff des semtu, fatum liegt nicht
darin. Der andere ausdruck ana semtisu illik, fato suo obiit.'- kann
' MDOG. 35, p. 43 n.
^ In der ncu gefundencn akk. vci^ion der charta des Xerxes von Persepolis wird
das ap. gadavd a^iyara durch ultu miihhi sa abnia dnriifius ina M-im-it el-li-i-ki uber-
setzt. l)amit ist die bedeutung „ sterben“ fur ,, vom throne geheri “ als euphemismus
fur den tod eines koniga festge&tellt.
DEE TOD DES KAMBYSES
595
dagegen nicht von selbstmord gebraucht werden. Der selbstmord
ist gerade nicht ein rniUu ana semti, sondern ein mitutu ina um Id
semti, ein ,, sterben am nicht bestimmten tage
Genau das sagt die akkadische iibersetzung von Behistun ; sie
gebraucht einen ausdruck, der fur das ap. hvdmrs'yus die bedeutimg
,,fato suo obiit “ ausschliesst : mitutu rammdnim miti ,, darauf
Kambyses einen plbtzlichen-tod einen selbst-gewahlten war er tot ’‘d
Der altpersische text.
amriyatd ist ,, er starb “d hcdmfyus * hatte ich friiher, AMI. v,
133, nach der analogic von Beh. hvdipasya gegeniiber NiR b, 15
hvaipasya als vrddhi-adjectiv aufgefasst, und darin ein determinativ-
compositum mit instrumentalem sinn des Jiva- gesehen, wie in
xvdhdta- „ durch sich selbst geschaffen und xvddivarsta- von
sich selbst gezeugt Daran nahin Schaeder. he., schweren anstoss
mit leichtem aber grundlosem spott. Denn wenn man vrddhi fiir
unwahrscheinlich halt, tritt Benvenistes erkliirung ein, Gramm. § 298,
die ich gern annehme. weil sie meine erkliirung erst recht bestiitigt :
Danach ist xvd- iiberhaupt instrumental von xva- und es bedeutet
litt. (la) mort par soi-meme “. Den instrumental driickt das elamische
durch sein sufihx -ma aus. Dies instrumentale „ tod-durch-sich-selbst "
ist der gegensatz zu dem ,, natiirlichen tod den man mit ausdriicken
wie „ seines todes sterben " meint. Also :
,, Darauf starb Kambyses durch selbstmord.”
So haben es auch die beiden iibersetzungen verstanden, und mit
recht.
Ein euphemismus liegt nicht vor, da das wort fiir sterben nicht
vermieden ist, und die bei Xerxes von Dareios' tod gebrauchte
wendung ,, er stieg vom thron ” erscheint bier nicht. * Der ausdruck
ist ganz kurz, nur zwei worte. und doch ist mehr gesagt als nur ., er
starb ■■ ; durch hvdmrsyus wird die todesart bestimiut. Man hatte
^ \\ ei&sbach hatte richtig, gewiss auf gnmd des akkadischen „ durch eigne hand
ubersetzt, denn in ramdiiivu liegt ctwas von „ wille " : es win! ,, eigen “ oder „ selbst
sofern ding odor person dem widen unterstehen, und es ist in kurze gleichbedeutend
mit ina qdt ranidniia, z.B. XiR. b 21 : daher ist die ein^chraiikung ,, unbeabsichtigt
kann gemcint sein die er in der anmerkung in rueksieht auf die nicht angezweifelte
Herodot-erzahliing macht, nicht notig. nicht einmal ganz richtig.
“ Ziir schreibung siehe AMI. hi, p. 113.
3 Cf. AMI. V, 133.
* Im mittelpersischen sagt man euphemistiseh vitnrtan ,, decedere ”, pn baxt
sutan ,, fato suo obire im neupersischen az hain (d.i. az miynn) raft ,, e medio
exces^it ”, alles genau wie im lateinischen, w'ie ich in hmbhek auf die bemerkungen
R. G. Kents, Language, ix, p. 42, erwahnen mochte.
596
E« HERZFEIjI^'“~“
erwarten konnen „ er starb plotzlieh, vorzeitig Aber auch als nur
zufalliger ware Kambyses’ tod ein tod ina um Id semtisu, gegen die
bestimmung des schicksals. Bei aller ahniicbkeit kann Jivdmrsyus
nicht zu der allgemeinen formel fato suo obire geboren, mit der es
Schulze vergleicht und iibersetzt, sondern muss ein ausdnick fiir den
geschichtlichen sonderfall sein. Anders als z.B. bei den Stoikern, die
darin den letzten und hochsten act menschlicher willensfreiheit sahen,
verurteilt der Orient den selbstmord. Ein beispiel aus assyrischer zeit
sind die angefiihrten worte Sargons, eines aus sasanidischer zeit wird
gleich folgen. Die iranisehe verdammung geht noch weiter ; auch der
tod ohne erhen — der fall des Kambyses — wird mit hollenstrafe
geahndet. Da dieser gedanke schon in Kambyses’ letzten worten bei
Herodot anklingt, ist die allgemeine verurteilung des selbstmordes
erst recht die anschauung der achaemenidischen zeit. Daher will
Dareios mit nicht mehr als zwei worten von Kambyses’ tod sprechen.
Auch ohne die beiden eindeutigen und iibereinstimmenden iiber-
setzungen muss man so urteilen. Denn von einem konig, der seinen
bruder und nachfolger ermordete, seine mutter in den tod trieb, seine
gattin durch brutalitat totete und sich dann, am rande des wahnsinns,
seines konigtums durch seinen reichsverweser beraubt sieht, zu sagen
,, er starb eines natiirhchen todes statt einfach ,, er starb konnte
nichts andres meinen als ,, in solcher lage starb er, so unwahrscheinlich
das klingt, eines natUrlichen todes “ : dann wollte die inschrift die
geriichte von der thatsache des selbstmordes durch leugnen aus der
welt schaffen. — Und wenn die worte bedeuteten ,, er starb seines
schicksalsgemassen todes so konnte das, da es nicht ,, er starb
als greis “ heissen kann, in dem beispiellosen fall auch nur als verhiillter
ausdruck fiir selbstmord aufgefasst werden. Aber diese auslegung ist
ebenso unzulassig, wie die als ,, zufalliger tod die Schulze mit seinem
,, recht verstanden, -npo <f)v(7€u)s, nicht rrpo potpa? “ halten will.
Denn damit wird immer der tragische schicksalsbegriff, auch der der
Nemesis, in die worte hineingetragen, der ausschhesslich griechisch
ist und in altpersischer gedankenwelt nicht erscheinen kann.
Endlich giebt es eine parallele zu dem inschriftlichen ausdruck
im Artaviraz-namak, i, 11, vom tod Alexanders d. Or. : x^-at-skast
6 dmaxv duudrist „ selbst-zerstort fuhr er in die holle Den anlass
zur sage vora selbstmord Alexanders miissen erzahlungen wie bei
Arrian, \'ii, 27, gegeben haben, wo Roxane Alexander von dieser
absicht abbringt. Die absicht wird dabei ahnlich dem geschichtlichen
selbstmord des Julian Apostata und des khalifen al-Hakim ausgemalt.
DER TOD DES KAMBYSES
597
Die im Orient um Alexander wuchernde sage hat den gedanken aus-
gesponnen, um dem erzketzer, der incarnation Ahrimans auch noch
schimpf und siinde des selbstmordes anzuhangen. Die worte des
Artaviraz-buches driicken dieselbe verdammung aus wie die der
inschrift, imd yvat-slcast ist das genaue gegenbild von hvdmrsyus?-
In R. M. Rilkes ,, Buck von der Armut und vom Tode “ steht :
„ 0 Herr, gib jedem seinen eignen Tod,
das Sterben, das aus jenem Leben geht,
darin er Liebe hatte, Sinn und Not.“
,, Denn wir sind nur die Schale und das Blatt.
Der grosse Tod, den jeder in sich hat,
das ist die Frucht, um die sich alles dreht.“
,, Denn dieses macht das Sterben fremd und schwer,
dass es nicht iinser Tod ist ; einer, der
uns endlich nimmt, nur weil wir keinen reiften —
Das griechische veos S’ dnoAXvd' ovtlv' dv deos steht
im gegensatz zum orientalischen ,, Des Menschen Leben wahret sie-
benzig Jahre ‘h Der tod nach einem solchen langen leben ist der tod
ana semtisw, der tragische tod ist ina um Id semtihi. Dies in der
griechischen welt nicht vorstellbare gegenteil vom dnodaveiv
Kara fj-oLpav ist haddsu emedu ramdnisu, mitutu ramdnisu. —
hvdmrsyus ist der nicht-unsere, nicht-gereifte, nicht-bestimmte, der
selbst-gegebene tod.
^ Ebenso spricht Eusebios, wenn er Pilatus’ selbstmord mit den kurzen worten
«i/ro^ov€UTi7? caurou cyevero berichtet, damit von Gottes strafe.
Bird-names in the Indian dialects
By E. H. Johnston
T he dictionaries of the modern Indian languages, whether scientific
or otherwise, are alike in omitting certain material which would
be of considerable value both to linguists and to those whose fortunes
take them to reside in the Indian countrywide. I refer to the names of
birds. There is, of course, good reason for the deficiency ; not merely
is interest in ornithology not as wfidespread among the educated classes
in India as it is in England, but only a few of the commonest birds
have names recognized over large areas ; while relatively few English-
men resident in India are capable of recording the names with the
accuracy required by scholarship. The suggestion I should like to put
forward is that concerted steps should be taken by the learned societies
in the different parts of India to catalogue the varjfing names of the
different kinds of birds, the materials to be compiled perhaps on the
admirable lines of Sir George Grierson's Bihar Peasant Life. Such a
survey could not avoid meeting with serious difficulties ; names vary
from area to area, and in some cases the same name is applied
to different birds in different districts. Detailed knowledge of the
names is mostly to be found amongst the shikaris and birdcatehers and
for waterbirds among the boating and fishing castes, and it seldom covers
birds w'hich are rare in any tract. Further, the recorder must
be able to identify the exact species in each case, and it is not easy to
find such persons nor has the less qualified observer as a rule any book
on which he can rely to help himself out.
The advantages and the nature of such a survey may be best
illustrated by a list which I drew up many years ago of the different
wild duck in the JIongh\T district of Bihar. These names, which I
collected for purely practical reasons, were not then noted by me in
a scientific transliteration, and after this lapse of time I am not prepared
to deny that there may be minor mistakes of spelling in it. The list
refers mainlv to Pargana Pharkia, a marshy area in the part of Mongffim
north of the Ganges famous for its ducks, geese, and other wmter-
birds ; and it covers all the varieties of duck which are to be found
there in any number, and should be compared with the names, spelt
in popular fashion, in F. Finn, The Waterfowl of India and Asia
(Calcutta, 1909), whose order I follow. The omissions are not without
600
E. H. JOHNSTON —
interest. Thus geese, which seem to be increasing in number there,
are divided nearly equally between the Grey Goose (Anser ferus) and
the Bar-headed Goose (A. indicus), whose different aspect is obvious
from some distance ; yet I never came across anyone who distinguished
between them in name, though Finn gives special names as known in
the adjacent district of Bhagalpur and in Nepal. The Dwarf Goose,
of which I only saw one specimen, naturally had no special name. The
Smew (Mergus albellus) frequents in most years a certain tank in the
Jamui Subdivision of Mongh3T, but equally has acquired no name.
Among duck proper the Mallard (Anas boscas) only comes occasionally
to this district, and is not known to the inhabitants as a separate
species ; and I omit the Buddy Sheldrake, for which the regular Hindi
word, caku'd, cakai, alone is used. My list is as follows : — -
The Bed-crested Pochard (Netta Rufina), dumar.
The Bed-headed Pochard (Nyroca ferina), usually known in England
as the Common Pochard ; the two sexes are entirely different in
appearance and the female is also known as the Dunbird. The male is
called Idlsar (Idlsir in Hindi), and the female arun (Sk. aruna). This
latter name is interesting, as it recalls the aluna of Asoka’s Delhi-
Topra Edicts, where it is mentioned next to the Buddy Sheldrake.
Arum, in fact, does not imply any genuinely reddish tint, as compounds
such as bhasmdruna show, and “ dun ” probably represents the sense
as well as any other equivalent. It would be useful for the identification
of the name in the Asoka Edicts to ascertain if this name is applied
to any other sort of bird elsewhere in India. Finn gives cheun as the
name in Nepal, which shows the kind of mistake recorders fall into ;
for this is really the wigeon, the females of the two species looking
much alike when dead.
The Tufted Pochard (Nyroca fidigula), dubdo ; this bird, as the name
implies, is a remarkable diver.
The White-eye (Nyroca ferruginea), known as kurid dubdo or majltd
(spelling of last syllable uncertain) ; in the adjacent Beguserai Sub-
division of Monghyr, where it is the commonest duck, it is called
maitd. This bird is also a great diver, and the name possibly goes
back to Sk. majj.
The Small Whistler (Dendrocycga javanica), sill, the regular Hindi
name ; the same name is given to the Large Whistler, D.fulva, which,
so far as I recollect, I only saw in the Jamui Subdivision.
The Comb-duck (Sarcidiornis melanonotus), naktd. Forbes
Hindustani Dicticmary gives it in this form also, and the name is
BIRD-NAMES IN THE INDIAN DIALECTS
601
presumably derived from nah, by reference to the curious growth on the
beak. The sexes separate in the cold weather ; I saw only females
in Pargana Pharkia, only males in the Jamui Subdi\dsion.
The Cotton-teal (Nettopiis coromandelianus), hahld.
The Wigeon {Mareca penelope), cihu.
The Shoveller {Spatula clypeala), sakcur.
The Pintail {Dafila acuta), dighoch (Sk. dlrghapuccha) ; Finn notes
this name from places as far apart as Nepal and Sind, but the regular
name in Hindi is sikpar (Prakrit, siihka. JRAS., 1934, 614).
The Spotted-bill (Anas poecilorhyncha), nadlm.
The Gadwall (Chaulelasmus streperus), mail.
The Garganey (Querqmdula circia), adhaiigi, presumably from
Sk. ardhdhga.
The Common Teal (Nettiiim crecca), gain.
Note
As regards the White-eye, Professor Turner suggests to me the
spelling majlthd and maltha and a derivation from Sk. wanjislha.
Sindhi having a similar doublet from the same word. The bird
could be described as madder-coloured, and I am prepared to agree
on both points.
VOL. Vin. PARTS 2 AND 3.
39
Note on the Ancient North-Western Prakrit
By Stex Koxow
rpHE discoveries in Chinese Turkestan have brought to light
extensive remains of one or two ancient Indian dialects. In
the first place we have the Dutreuil de Rhins manuscript, which has
found its way to Paris and Leningrad, and, in the second, the numerous
Kharosthi documents found by Sir Aurel Stein and now available
in the splendid edition of Messrs. Boyer. Rapson. and Senart. I shall
distinguish them as Dhp. and Doc. respectively.
In Dhp. we have before us a remarkably consistent dialect, an
ancient Prakrit, which must have been reduced to writing at a com-
paratively early date and used by some Buddhist school in religious
literature. In my edition of the Indian Kharosthi inscriptions
I have tried to show that it is essentially identical with the form of
speech used in those records, over a large area, where we. at the present
day, find Dardic languages, Lahnda, and Sindhi. and I have also
drawn attention to the existence of traces of other Buddhist works
in the dialect. It is a 'priori likely that Dhp. has been translated from
an older version in a different dialect, and it is easy to show that the
translators were also acquainted with the Sanskrit forms of several
words. But it is a likely h^'pothesis that the translation was made
at a time when Sanskrit had not yet come into general use as the
sacred language of the leading Buddhists of North-Western India.
The Doc. language is far less consistent, and it has undergone so
many modifications that we must necessarily assume that it had, for
some not quite inconsiderable time, led an independent existence of
its own. partly under the influence of non-Indian speeches. AVe can
see that a Buddhist literature in Sanskrit had begun to come into
vogue, at least towards the end of the period covered by these records,
i.e. the end of the second century.
In a short note. JA., x, xi.x, p. 411. Senart wrote that “ les tablettes
rapportees par le Dr. Stein, bien qu'elles proviennent a peu pres de
la meme region, ne presentent pas les particularites phonetiques qui
characterisent le manuscript Dutreuil de Rhins . I have never
been able to accept that view. So far as I can see, we have before us
two forms of one and the same dialect, one normalized as a literary
604
S. KONOW —
language, the other developed in a foreign country by the descendants
of an ancient Indian colony, under the influence of more than
one alien tongue.
As long as the Leningrad portion of Dhp. is inaccessible, it is not
possible to arrive at certain results about all details. I have repeatedly
tried to get reproductions, but always in vain. Our conclusions must,
therefore, be based on the excellent plates accompanying Senart’s
masterly edition of the Paris folios. And, as a matter of fact, they
are sufflcient in order to enable us to settle the question about the
home of the dialect with some confidence. That has been done in
an excellent essay by Jules Bloch, JA., x, xtx, pp. 331 If., at the
hand of certain phonetical features, notably the treatment of compounds
consisting of a nasal and a stop. We must, it is true, take exception
to his statements that hg, ngh occur as g, gh, and nd as n, respectively.
We find, it is true, g < hg in sagama < samgrdma and Icadigara, Pali
kalihgara, but in sagama g is initial in the second part of the compound
word, and with regard to kadigara it is possible to compare Skr.
kadahkara, Panini, v, i, 69. Everywhere else g < hg is provided
with a hook above or below, so that we can hardly have to do with
an ordinary g. And for nd we always find n and not n} Bloch s
main results are, however, unobjectionable : “ la seule region qui
admette a la fois tons les traitements characteristiques du dialecte
dans lequel sont rediges les fragments Dutreuil de Rhins est celle du
penjabi occidental et des parlers montagnards du Nord-Ouest. On
est done en droit d'affirmer que I'original de ce manuscrit, s’il a ete
ecrit dans ITnde, n‘a pu I'etre que dans cette region. II est egalement
possible, bien entendu, qu'il I'ait ete plus au Nord, en un pays ou
Ton aurait parle aux premiers siecles de notre ere un dialecte indien
disparu depuis, et formant avec les parlers de I'Extreme Nord-Ouest
un groupe linguistiquement continu.’’
In my edition of the Indian KharosthI inscriptions I have drawn
attention to several details where we find parallels in Dardic languages.
In the present paper I shall try to show that there are many important
points of such agreement, and that this remark also holds good with
reference to Doc. The great chronological gap of more than a thousand
years between Dhp. and Doc. on one side, and Dardic on the other,
prevents us from expecting a thorough agreement.
Grierson, JRAS, 1925, pp. 226 If., has shown that the combination
^ See my remarks Festschrift fur Ernst Windisch, pp. 85 ff. There are several
misprints in this paper, owing to the fact that proofs could not be sent to me.
NOTE ON THE ANCIENT NORTH-WESTERN PRAKRIT
605
vowel -f a geminate consonant is treated in different ways in Indo-
Aryan vernaculars. Thus Prakrit bhatta < bhakta becomes bhatt in
Panjabi, bJiat in Gujarati and the literary languages east and south of
the Panjab, but bhat, bat in Sindhi and Dardic. The common view
is that the last-mentioned forms have passed through the Panjabi
stage, but Grierson maintained that the geminate consonants did not
exist in the Prakrit of the North-West. In Dhp. and Doc. long
vowels and consonants are not generally marked as such, so that
bhat a might be bhata, bhatta, or bhdta. But bhata is excluded because
the syllable is long, and bhdta because -I- regularly becomes -d-, at
least in Dhp.'^ We can only think of bhatta. If those scholars who
maintain that the t of Sindhi and Dardic is derived from tt are right,
this feature does not prove anything about the position of the Dhp.
Doc. language.
The treatment of the r- vowel varies in Dhp. and Doc. We fre-
quently find the usual change to a, i, or «, e.g. Dhp. svadi < smrti,
kida < krta, niucu < mrtyu ; Doc. kada < krta, kisi < krsi, muta <
mrta. In other cases the r-element is preserved ; cf. Dhp. pradhai'i,
padhavi < prthirl, dridha <drdha, prudhijana < prthagjana ; Doc.
krida < krta, ghrita < ghrta, prichati, pruchati < prcchati. Turner ^
has compared prichati with Kashmiri ‘‘ asks ” and drawn atten-
tion to the development r > n in most Dardic languages. Doc. is
here, on the whole, more in accordance with Dardic than Dhp., which
was a literary language, probably to some extent influenced by the
Prakrit from which it was translated.
With regard to stops the general rule in Dhp. is that they are
preserved as initials and voiced as intervocalic. Forms such as loka
with k can hardly represent the actual pronunciation, the less so
because -k- occasionally also stands for -g-, e.g. in nadakara < naddgdra,
parakata < pdragata. A similar confusion is not rarely met with
in Doc., e.g. in citughi and jit ughi ; tamda, tamta and damda ; poga
for bhoga. We cannot draw the inference that voiced stops were
liable to lose their voice. Such writings find their explanation in the
voiced pronunciation of voiceless intervocalic stops, and partly probably
also in the desire to avoid the pronunciation as fricatives, to which,
e.g. the g in poga seems to bear witness. I shall not, however, try
to analyse such details which seem to point to the existence of fricatives
^ Senart and those who have copied his text have failed to distinguish between
the ak?aras ia and da. Also in Doc. it is often difficult to choose between t and d.
- Gypsy Lore Society, Monographs, Xo. 4, p. 7.
606
S. KOXOAV —
especially in Doc., because it would take too much of the limited space
at my disposal.
There is an evident tendency to deaspirate voiced aspirates, a
tendency which manifests itself in spite of the literary tradition, in
which the influence of other Prakrit.s may have played a certain role.
Even in Dhp. we find forms such as jai < dhyayan, clunadi < dhundti,
and conversely drnyha <i durgdt, ghadkedi <_ ghatayciti, kusidhii <
liusida. samkUta < svdmitd, saghadhadhama < samskrtadharma, which
point to a pronunciation of d/i as d, if we have not here indications
of fricative sounds. In Doc. the examples are numerous ; cf. ghuta
and grida < ghrta. vigu ^ < vighna. sigra and hghra < slghra. jdna
dkydna, ajesana < adhyemnd, guta < gudha, daridavo and dhatidaio
< dhdrayitavya, tarmena and dhannena < dhannena. anusaniti <
(DiHsamdhi, bhagena and bagena < bhdgena. poga < bkoga. durlapa
< durlabka, etc., and convensely, dhada, damda and (amda < dnnda,
dhadiwu and dadima < dadima, dhasami and dakimi < dasame,
etc.'^ It is difficult to account for this state of things, otherwise than
by assuming a strong tendency towards deaspiration, and it is of inteiest
to recall the fact that such deaspiration is a prominent feature in
Dardic, though even there it has not become the rule.®
Before leaving the subject of uncompounded stops it will be
convenient to mention one curious feature, viz. the occasional change
of intervocalic dentals to s or r. In Dhp. I have only found sagha
sndha/Hf.vi corresponding to Pali sainkhatadhamntdncon Cro II and
masHrit < mculhnmm B 11. Senart looked on both forms as mistakes,
but thev find some confirmation in Doc. asDnatra, npiiudHi, ujhitnaUn
<adhinidtra, and tiat'iu < maiUin We may also compaic hiilo
<kthila, whicli is common to Dlip. and Doc., because it ma} haM
passed through s/.sdu. in tlie same way as sdsana occurs as kimm
in Dhp. and saktna in Doc.
This change of a dental to a or 2 must have been found in the
dialect from which KhotanI Saka and the so-called Tokharian ha^e
^ Tlie ^igniti< atioii ()f tho .stroke above (j i.s not of interest in tins eonnection.
2 Some of the instances of rlh for d may point to a fricative pronunciation, and m
the Khotan document Xo. GUI, which has several instance.s of Iranian features, ic
regular initial dh < ,/ i.s almost certainly a fricative. With regard to forms such as
tamdri < dnnda it .should be borne in mind that it is often very dilhcult to c loo.sc
between t and d. . .
^ Cf., e.g., Morgenstierne, Ih part on a Lhigiii.-<ttc Mi-^sion to Afghanistan, p. .)! .
Report on a Lingii istic M tsnion to S ot th-\\ extern. India, p]i. ZD, 41>, .iS.
‘The true explanation of thus word wa.s given hy [aiders, .S’B*4II., 1933,
pp. 1000 ff.
NOTE ON THE ANCIENT NORTH-WESTERN PRAKRIT
607
borrowed some of their Indian loanwords, for we find Saka vT/ysn,
" Tokharian '' slm < slta. Saka mnysila < nnlhila.
The Saka forms and some of those found in Doc. seem to point
to r rather than s, while Dlip. .s- may represent an attempt at noting
the voiced s. In that case it would be possible to think of a voiced
dental fricative as the intermediate stage. At all events there cannot
be much doubt that Doc. nmnu " wine " is the same word as Kashmiri
to which word i\Iorgenstierne has drawn mv attention. M<ts
can well be an old maz with the usual devoicing. and. on the whole, it
is tempting to compare what Grierson ^ calls the Zetacism in Dardic.
The form sisila mentioned above has a parallel in Ashkun cicila.
Kati ceil " soft k
As shown in the paper quoted above, (1. 54). Dhp. distingui.shes >i
and n in the same way as a large group of Indo-Arvan vernaculars
compri.sing Sindhi, Lahnda, Panjabi, Raja.sthani, Gujarati, and
Marathi.^ In Indian KharosthI inscriptions there is no consistency,
but we are left with the impression that the two sounds tended to
coalesce in the dental h. Such is almost certainly the ca.se in Doc., as
stated by Rapson in the edition mentioned above, p. 605.
According to Grienson. l.c., para. 117. "in Dardic the e.xistence
generally of the n-sound is uncertain. It certainly does not e.xist in
Ka.smirl, and has not been noted elsewhere, e.xcept in Kafiri. where it
may be due to contamination with the neighbouring PSto. and in
Sina, where it is a secondary sound, arising from the pro.ximity of
another cerebral sound, as in the word Sinii itself.”
Here there is accordingly an apparent difl’erence between the system
of Dhp. and that of Doc. and. so far as we can see. Indian KharosthI
inscriptions. It is. however, remarkable that the Kurraiu casket
in.scrqition, which contains a quotation of a canonical passage written
in practically the same language as Dhp.. ha.s no trace of the Dhp.
distinction between >i and ». We are left with the impression that
Dhp. in this respect represents a normalization which may be due to
the influence of another literary Prakrit, or belongs to a limited
territory within the area covered by the dialect, where the treatment
of n was different.
In Indian vernaculars an intervocalic ih usuallv becomes r. In
Dardic, however, it remains, or the change is at least rare and. according
to Grierson, l.c.. p. 101. in such eases probablv due to borrowing. It
’ On the modern JixJo-Aryan Laniriiage^, jiara.
- Cf., c.g., Bloth, La fonnatiun d* la himjnf Mmaiht, jiara. 132.
608
S. KONOW
is, on the whole, a comparatively late change, so that nothing can be
inferred from its apparent absence in Dhp. and Doc. Forms such
as jyhT^. jinaviva < prnam iva can hardly count, because the accusative
termination had long ceased to be am. On the other hand, the opposite
change of -v- to -m- in Dhp. emmneva < evameva, sabhamu < sam-
bhavam, namo < ndvam, etc., Doc. emit < evam, cimara < clvara.
gamesati < gavemte, etc., is isolated and difficult to explain, just as
the corresponding change in Kashmiri namath ninety Torwali
noyn “ nine ”.
The Dhp. treatment of nasals followed by mutes has, as mentioned
above, been examined by Bloch, and I shall only add that the change
of nd to n is occasionally met with in Dardic, e.g. in Kashmiri ra«“ <
randa. In Doc. there is great inconsistency. We find upasainb.
ttpasamg, and upasamgh < iipasafik, dramga < dranga, karncuh, and
kamjidi, khamja but gamJiavara, damda, chimnati, and chinita, bimnanti,
and bimnita, etc. If we bear in mind how often we ourselves believe
that we pronounce letters which are no more heard, we get inclined
to lay special stress on such writings which agree with the state of
things in Dhp.
In connection with nasals it may be of interest to consider the
treatment of old tm. It regularly becomes tv, e.g. in Dhp. and Doc.
atvan- < dtman-. Writings such as Dhp. iitvari, Pali uttarim, Doc.
ntvaravarsi < uttaravarse, however, show that the v was hardly
sounded. If the common Doc. word tanu “ own ” is derived from
dtman- and not from tanu-, it is of interest that it has its exact parallel
in Dardic, cf. Turner, Monograph, p. 14.
Both Dhp. and Doc. preserve old y and do not change it to j like
most Prakrits ; cf. Dhp. yujatha. Doc. yudhammi, etc. The same is
the case in most Dardic dialects.^ Grierson ^ remarks that ny becomes
n in Kashmiri, but not elsewhere in India, and similarly we have
ana < anya, etc., both in Dhp. and in Doc. The difference in the
treatment of y is old, and therefore of some importance.
Of interest are also the numerous compounds where r has been
preserved, because we know that also in this case the different treat-
ment can be traced back to the third century B.c. The writing is not
consistent, but forms such as Dhp. krodha, gradhadi, tnhi, drumapatra.
prldipramoja, bramana. Doc. kratna, grahita, triti, dramga, brammana,
bhrada, etc., clearly show the prevailing tendency. Among modern
‘ Cf. Turner, BSOS., iii. p. 208.
“ LSI., viii, ii, p. 245.
NOTE ON THE ANCIENT NORTH-WESTEEN PRAKRIT
609
languages Dardic and, to some extent, Lahnda, SindhI, and Western
Pahari show the same state of thingsd
The change of p to 6 is apparently comparatively late in Indo-
Aryan vernaculars. We cannot, therefore, lay much stress on the
fact that it is unknown in Dhp and Doc., if we abstract from b <v
in particles after old »i, e.g. in Dhp. siha ba < simham iva. It is.
however, of some interest that the change is also absent in Kashmiri,
while some other Dardic languages often, but not always, have it.-
Attention may also be drawn to the fact that tv and dv are usually
preserved in writing, e.g. in Dhp. catvari, salvana, vulva, etc., Doc.
satva, dvara, etc. Forms such as Doc. caparisa < catvdrimsat, bddasa <
dvddasa, biti side by side with dv-iti < dvitlya, etc., are, therefore-
suspect of borrowing. As mentioned under the head of tm the v
was probably not sounded, cf. also Dhp. iidhvaradha < uddharata.
The change of tv, dv to t, d respectively, which is already found in
the North-Western Asoka dialects, is common in the Dard group.®
It is also tempting to compare the gerund termination ti in Dhp.
baheti, pram-ajeti, Doc. aprochiti. pahyiti, etc., which is already found
in the North-Western Asoka edicts,^ with Kashmiri ith, Kati ti.
An outstanding feature is the preservation of, and careful dis-
tinction between, the three old sibilants, s, s, and s, which again have
their parallel in the North-Western Asoka dialects and in Dardic.
The state of affairs is mainly the same in Dhp. and in Doc., even in
minor, but significant details. They both also agree in sometimes
using Sanskrit forms side by side with the genuine ones, but that
is a feature which is too well known in most Indian languages to neces-
.sitate a discussion of details.
There is only one feature which cannot be traced in Dhp.. \dz.
the tendency to voice intervocalic sibilants like other intervocalic
consonants ; cf. Doc. ija < isa, kojahja < kausalya, dajha and
da.sa < ddsa, tivajha < divasa, etc. We cannot a priori decide whether
the apparent absence of all such forms in Dhp. corresponds to the
actual state of things or is due to the absence of letters for voiced
sibilants. But it is of interest that the same tendency is found in
Dardic,® e.g. in Sina bazodu < vasanta, mdz < mdsa.
^ Cf. Turner, Monograph, p. 21.
~ Cf. Grierson, Vernucitlars, para. 367.
^ Cf. Turner, Monograph, p. 14 ; but Prasun cipu, Katl ctca, Waigeli cata “ four
^ Cf. Thomas, Acta Orientalia, xiii, p. 78.
^ Cf. Turner, BSOS. v, p, 130.
610
S. KOXOW —
Though old s is generally preserved as s, there is one curious exception .
For Sanskrit SMct “ pure " Dhp. has always si/^i and Doc. once
suci. IMorgenstierne has drawn my attention to Kati suc,^ an exclama-
tion used in sacrifice. Have we to do with an old dissimilation 1
Dhp. kisana, Doc. sasana < sascuui, and perhaps sikla < * sisiln
< ktJnla. are examples of the opposite tendency.
Writings such as Dhp. mmana. Doc. sanuina side by side with
the San.skrit form sminana show that sr regularly became s just as in
Dardic and Kati.^
Old h- occurs as s and as sp ; cf. Dhp. asa < ah-a, vispa < visva,
Doc. aspa, vispa, etc. Writings such as Doc. esvari < aisvarya,
svana < svuna are evidently “ learned ". The usual forms can be
compared, e.g. with Sina sS, Kashmiri /<««“ ” dog ‘"j Sina aspo
"horse”, etc.
AVith regard to s it is a notable feature that the numeral “ six
which begins with ch in all Prakrits except the North-AVestern, and in
all Indo-Arvan vernaculars except Dardic, has preserved the initial
s in Dhp. Doc. so. Also the treatment of the compound ks is the same
as in Dardic. There are, it is true, in Dhp. and Doc. as everpvhere
else, some stray examples of the change to kh, but as a general rule
we find an aksara which resembles ch. but differs from it in not con-
taining the cross-bar, and which had been variously transliterated
as As and as ch. It cvidentlvhad the same sound as the Dardic cerebral
affricate sound ch mentioned by Turner, Monograph, p. 11. Cf. Dhp.
cheuja < ksaya. cnchutna (beside cakhunta) < cak-pimn, Doc. chitra
< ksitra. rachidam < raksitavyani, etc.
If we turn to the dental s, we are met with some difficulties. The
sr in Dhp. visravatena < visravatu and the .s', which means the same
thini;. in saga < sahgn are peculiar. Is it possible to compare the
curious r in Kashmiri scuts'^ " pure ” ?
The change of ts to ts in Dhp. bhetsidi < hhetsgati. malsam >nat-
sgundm. Doc. sameutmra ■- satnvulsura is not much different from the
usual treatment in other Prakrits. More peculiar is the development
of ins to tnts in Dhp. ahilki < ahiinsd, satsara < sainsdra, Doc.
marntsa < nidnisa. etc. In Dhp. we also find prakijadi < prasatn-
santi, where the ts sound has become voiced. So far as I know a
^ Cf. Robertson, Thf Kafirs of the liindu-Kvsh, p, 421.
^ Cf., e.g.. Morgen&tierne, UeportmAJijhnmstnn., p. o-S.
NOTE ON THE ANCIENT NORTH-WESTERN PRAKRIT 611
similar development has only been traced in SindhI. where we find
haiiju < hamsa}
Old sv occurs as sv and as s ; cf. Dhp. svuya < svarya, salavhu
< svaldbham. rasa < hrasva. Doc. svaya < svayant, etc. In Doc. we
occasionally find sv instead in svasti and regularly in svasu < svasr.
It is tempting to compare Torwali su. Garwl ispo and perhaps Khowar
isp}isdr, and it is perhaps possible to think of a kind of dissimilation.
The word hrisn is also of interest in another respect, as showing that
the word for sister " was the same as in Dardic, where only Kashmiri,
which has been largely influenced by the Midland languages, has
adopted bene.
With regard to the compound st the rule is that it remains between
vowels and is changed to th when initial ; cf. Dhp. hnsta < hasta.
thala < slhala, Doc. hasta < hasta, thana < stlidna. There are, how-
ever, also instances of initial st, e.g. Doc. stita < sthita, stora “ horse "
stri and istrl < strj. A glance at the material registered by Grierson,
Vvrnacukirs, para. 290. will show that also here there is a remarkable
agreement with the state of things in Dardic.
In making these remarks I have not attempted to give an
exhaustive sketch of the phonetical system of Dhp. and Doc. But
I have tried to draw attention to some important features which,
in my opinion, make it absolutely certain that Dhp, and Doc. are not
two different languages, but clo.sely connected forms of one and the
same ancient dialect. ^loreover. there are so many essential points
of agreement with Dardic that it is diflicult to avoid the conclusion
that also Dardic is derived from a Prakrit of the same kind as that of
Dhp. and Doc. The cerebral h in Dhp. and the change of n\s to
mts in Dhp. and Doc., it is true, point to a territory where wo, at tlie
present day, do not find Dardic speeches. Grierson has. however,
repeatedly drawn attention to Dardic trait.s in modern vernaculars
outside the Dardic territory, and also the find-places of Kharosfhl
inscriptions written in the old Xorth-Western Prakrit rai.se a strong
presumption in favour of a.ssuming that this Prakrit was subsequently
ousted from a large tract of country, where we at the present day find
languages such as Lahnda and SindhI.
Dhp. shows that the North-Western Prakrit was, at a comparatively
earlv date, used for literary purpo.se.s by a Buddhist sect. And the
remarkable consistency apparent in the Dutreuil de Rhins manuscript
* Cf. ’■aiijhti, Panjabi vai'ijh < vainki, and Doc. ^ainmyn < natnsaya.
612
NOTE ON THE ANCIENT NOKTH-WESTEKN PRAKRIT
shows that this was done with great skill and insight. We cannot
tell how the development would have been, if this literary activity had
been continued. But we know that the leading Buddhist school of
the North-west soon gave up the use of Prakrit and adopted Sanskrit
as their sacred language. We thus understand how these parts of
the area of the ancient Prakrit where the literary activity was continued,
in Sanskrit, came under an increasing influence of the Midland. Such
has especially been the case in Kashmir, though even there we still
find remarkable features which remind us of the Prakrit of Dhp.
and Doc.
The Military Colonization of the Caucasus and
Armenia under the Sassanids
By J. H. Kramers
rriHE ancient Arabic historian al-Baladurl (d. 892) begins his chapter
-L on the conquest of Armenia by a description of the political
conditions of those regions in Sassanian times. According to the local
historical tradition, obtained from inhabitants of several Armenian
towns, there had been a time when the people of the Hazars in Southern
Russia were making continuous raids over the Caucasus passes and
penetrated Persia as far as al-Dlnawar in Media. The first king to
take energetic measures against these raids was Kubad (Kawad,
488-531). One of his generals ravaged Arran (Albania) between the
Araxes and the Kura ; then Kubad came himself and founded or.
better, fortified in this region the towns of al-Baylakan, Barda'a,
and Kabala. He erected also a wall of brick which extended from the
country of Sirwan in the east as far as the pass called Bab al-Lan,
the “ Pass of the Alans His work was completed by his son Kisra
Anusirwan (Husraw I, 531-579), who fortified farther to the north
the towns of al-Sabiran and Maskat,^ and finally the very strong
town of al-Bab waT-Abwab, on the site of the later Derbend. The name
of this town, “ the Gate and the Gates,” is explained by the fact that
its fortifications comprised the gates to several mountain passes.
Here the text of al-Baladurl (ed. de Goeje, Lugd. Bat., 1866, p. 194)
continues as follows : — LJl C.^ oAa ^ G
“ He made dwell in these places which he had buUt a kind of people
whom he called al-SiyasIgIn ”. Continuing the same tradition
al-Baladurl describes the conquests of Anusirwan in the western
direction, in Georgia, as far as the Black Sea, and in the south-western
direction in Roman Armenia. Here were conquered in the first place
Dabll (Dwln) and Nasawa (Nahcewan), and further the fortress of
Wayas “ and several fortresses in the country of al-Sisagan. Here
(p. 195) the text continues :
1 On the topography of all these places in Albania cf. Marquart, Erdnsahr,
pp. Ill, 118, and the map accompanying W. E. D. Allen’s History of the Georgian.
Pfople, London, 1932.
- Arm. Tfiyoc, cf. Hubschniann, Idg. Forsch., xvi, p. 469.
614
J. H. KRAMERS
“ He made dwell in these fortresses and
strongholds strong and valorous men from Siyasiglya Finally
we read, at the end of the section on the conquests and the reign
e.
of the Persians in those regions (p. 197) : 1 y Jjj"
‘C-q y L. ^ “Armenia
continued to be dominated by the Persians, until the appearance
of Islam ; many of the Siyasigin then left their strongholds and
their towns, which consequently were ruined, while the Hazars and
the Eomaus recovered the territory they had originally possessed
The orthographies (>=>^1x^1 and .L-, were adopted by
de Goeje partly on the base of the different and generally
unpunctuated readings of his manuscripts (see below) and partly on
the assumption that the word must be related to the name of the
Armenian district of al-Sisagan and that it denotes the inhabitants of
that region. In the note on p. 194 of his Baladurl edition de vioeje
says ; “ Est populus cujus genealogiae princeps appellatur Sisag,"
while referring to St. JIartin, Memoires sur VArnienie, Paris, 1818, i,
pp. 207-214. Here St. Jlartin discusses the text of Moses of Khoren's
History (book ii, cli. 7) on the province of Sisakan, which is the north-
westernmost pro\'ince of Great Armenia, hung betw'een the Araxes
and the Lake of Sewan and bordering on Albania ; the older
.Armenian name is Siounik'. Moses of Khoren derives the name
Sisakan from a heros eponywos Sisak ; this Sisak is, however, as
Hubschmann also {Idg. Forschnnyen, x^^, p. 263) thinks, only an
imaginary forefather, whose name was deduced from the form Sisakan.
Now the rea.son of de Goeje's assumption can be no other than the
fact that in the second passage quoted from al-Baladurl (p. 195),
the were placed also in al-Sisagan.
I do not know if de Goeje is the first to have made this identification.
For in Thornberg's edition of the Chronicle of I bn al-Atlr there occurs
a parallel to al-Baladuri's fir.st passage in tome i, p. 319 (edited 1851).
where the word in question is read although the MS.
readings do not seem at all to make such a spelling more probable
than any other. After de Goeje, however, the identification given by
him has never been questioned ; it was adopted by Marquart in his
MILITARY COLONIZATION OF THE CAUCASUS AND ARMENIA 615
earlier works {Osteurop. und Ostasiat. Streifziige, Leipzig. 1903,
pp. 37 sqq. ; Erdnmhr, Berlin, 1901. p. 120) and by Hubschmann {Idq.
Forsch., xvi, loc. cit.).
Parallels to al-Baladurl's first pas.sage are found, besides in Ibn
al-ALr, also in Kudama (ed. de Goeje in EGA. vi. 1889), p. 259 —
where the reading M-as adopted — and in the geographical
dictionary of Yakut (ed. Wiistenfeld, i, p. 221)— where the edition
gives the same spelling as Ibn al-Atlr. A parallel to al-Baladuri s
second passage (p. 195) is found in Ibn al-Fakih (ed. de Goeje in
EGA. V, 1885), p. 288, where de Goeje has printed in
which the addition of the possessive ending gives in any case a
better reading.
Ibn al-PakIh has, moreover, a passage, to which a parallel is not
found in al-Baladurl. It is found on p. 291 of de Goeje's edition in a
description of the fortification of the town of al-Biil) wa'1-Abwab
by Anusirwan and of the wall extending from this town to the
mountains over a distance of seven for.sahs. Here we read :
AS Aj A.* 'ALLw* ^ 1 1 oAA ^
^ jLi Cy “He
made in this distance of seven farsahs seven passages ; each one of
these was dominated by a town, in which he had placed Persian
warriors named al-SiyasikIn A parallel text is found in Yakut,
i, p. 440, where there is printed The latter reading
is made also much more likely by the MSS. of Ibn al-FakIh, but
de Goeje, by his Sisagian or Sisakian theory, has again adopted a
reading complying with that theory.
Finally the same people are mentioned probablv in al-Mas‘udrs
3[itrug al-Dahab, Paris edition, ii, p. 75, where it is said that they used
the so-called Siyawardi battle-axes. It is true that the Paris edition
calls them -11. but Marquardt {Streifziige, p. 37) has pointed
out that the Cairo edition, p. 89, has here ‘w-^a—LUI .
Now a comparison of the different MS. readings ^ puts it bevond
question that the unpunctuated original readings of the three text
passages of al-Baladuri and their parallels is and
•u=>= — while the passage of Ibn al-Fakih, p. 291. and its Yakut
' For footnote, see p. 616.
616
J. H. KEAMEBS
parallel go back to My conclusion is that the first group
must be punctuated (and — which is, indeed, the
reading of the British Museum MS. of Ibn al-AtIr — and the second
— ^which is the printed reading of Yakut, i, p. 440. The
word would render then an arabicized plural of middle Persian
nisdstag, belonging to the middle Persian verb nisdstnn, the causative
form of nisastan (cf. H. S. Nyberg, Hiljshuch des Pehlevi, ii, Glossar,
Upsala, 1931, p. 161). The meaning would be “ somebody who has
been made to dwell in a certain place ” and, in a pregnant sense, a
“ garrisoned warrior ”. The Arabic verb used in the first
passage of al-Baladuri is the exact counterpart of nisastan in this
sense.
This interpretation is much more obvious than that of Sisakians,
because an Arabic plural derived from would yield
and the addition of the ending -iin or -In would be
abnormal in the highest degree. And as to the historical facts, it
appears from al-Baladuri, that those garrisons were firstly laid in
Albania and the Caucasian towns, and only afterwards in Armenia,
amongst others in al-Sisagan. Moreover, the text of Ibn al-Fakih,
p. 291, states expressly that the garrisons consisted of Persians.
Another interpretation of the word concerned is incidentally given
by J. Markwart in his paper, “ Np. aSina Freitag " {Ungariscke
Bibliotheh, i, 13), p. 83,’- where, without referring to the Arabic
Footnote to p. 613.
al-Baladuri. p. 191
(aUMSS.)
Ibn al-Fakih
Kudama p. 257.
Ibn al-Atir i, 319.
p. 193
p. 195.
(two MSS.)
(one MS.)
^ 1 ^. 1 .. 1 .... 1 ^
p. 288
p. 291.
JALL-VI
k.. V I
^ I owe this reference to the kindness of Profes.sor V. Minorsky.
MILITARY COLONIZATION OF THE CAUCASUS AND ARMENIA 617
historians and geographers cited above, he takes it for granted that
the Sasanian military colonists in Daghestan were called spdsigdn
Markwart here interprets the word as
“ Dienstleute ”, deriving it from Pahlavi spds, which, as appears from
the Armenian loanword spas, may have had also in middle-Persian
the meaning '' service " (cf. Nyberg, Hilfsbuch, ii, p. 205). I cannot
agree with this explanation ; (1) because a form spdslk has not been
actually proved to exist, while in Persian sipdsl apparently means
only a “ beggar ”, according to the Burhdn-i Kdti‘ ; (2) because the
majority of the manuscript readings is in favour of the reading ;
(3) because al-Baladuri's text gives a kind of interpretation by the verb
jjd'. Moreover, the form — . does not deserve any regard.
because the texts prove that the final b is the suffix of the 3rd person
preceded by the oblique form of the masculine Arabic plural.
The information given by the Arabic authors on these conquests
reposes on local traditions gathered in early Islamic times and does
not occur in this form in the Arabic versions of the Persian Royal
Annals, the Hwatdy-vdnmk. It is not found in the history of Sassanian
Persia by al-Tabari, nor in Firdawsl's Sdhndma. For this reason the
corrupted reading in the Arabic texts cannot be ascribed to the mis-
interpretation of a Pahlavi original. The two different original forms
and — with a prothetic vowel as in
— also point to an endeavour of the early Islamic collectors of
local traditions to render a word really heard.
From an Arabic granunatical point of view the ending -In — used
by Ibn al-Fakih even in the nominative — seems to be a compromise
between the Persian ending -an and the Arabic pluralis sanus ending
in -un (gen. -In). I do not know another instance of the plural of a
Persian word being made in this way, but it certainly can be brought
under the rule that the proper names of men form their plurals in
this way.i The plural of non-Arabic words is formed by preference bv
a so-called “ broken plural ” (cf. Siddiqi, Studien iiber die persischen
Fremdworter im klassischen Arabisch, Gottingen, 1919, p. 20 sqq.) ;
only the ««s6a-forms — which often represent as well the Persian
ending -i{k) — have generally -lyun.
Further, it is to be noticed that the causative verb nisdstan is
* Wright's Grammar of the Arabic Language, 3rd ed., i, p. 195.
VOL. VIII. P.CRTS 2 AXD 3.
40
618 MILITAKY COLONIZATION OF THE CAUCASUS AND AHMENIA
essentially middle-Persian, for which modern-Persian has nisandan
or nisdhtan. In literary Pahlavi it is used in a cognate sense in the
Catalogue of the 'provincial Capitals of Erdnsahr by Markwart (ed.
Messina, Eome, 1931), pp. 9, 10, 17, for the founding of fire temples
or a camp, while in Manichaean middle-Persian we find it used for the
founding of Manichaean monasteries called mdmstdn (Andreas-Henning,
Mitteliranische Manichaica aus Chinesisch-Turkestan, ii, Berlin, 1933,
p. 11 [302]). Dr. Bailey kindly draws my attention to two passages inthe
Bundahisn (ed. Anklesaria), where nisdstak has equally the meaning of
"settled’’, used of men. In the first passage (pp. 106^’®) it is
said that nine of the primordial kinds or races of men passed the
sea FraUiart from X^vaniras and were settled in the other six
Karsvars (o dn 6 kisvar i ditlgar vitdrt u od nisdstak kart). Inthe other
passage (pp. 108^^, sqq.) it is told how the people of the Zangik, who had
sprung up under Azi Dahak’s reign from the mixture of men and female
demons, fled from Eransahr when came the reign of Freton, and were
made to dwell on the border of the sea (pat kandrak-i zrdy nisdstak
kart). The reading of nisdstak (p. 109^) is here corrupt, but as
Dr. Bailey informs me, the shorter text has here the ideogram
for “ to sit ”. The language of FirdawsI has, just in the
same meaning of populating a town — which is the meaning of
jUcl-UjJl — the verbs (ed. Turner Macan, p. 1379, with
relation to the foundation of Gor — nuitten erroneously Zor — by
Ardasir I) or jjJLiJ (ed. Mohl, vi, p. 214, in connection with the
populating of the town of Zeb-i Husraw with Roman prisoners by
Anusirwan).
We have to finish by pointing to the fact that the interpretation
given above of the Arabic texts is not without importance for our
knowledge of the military administration of the Sassanid Empire,
as it reveals the establishment of a system of frontier garrisons which
reminds us very much of the organization of military themes in the
Byzantine Empire in the seventh century and, in many ways, of the
frontier-posts called ribdt in the Islamic empire of the Caliphs.
Mala vihara
Par f Sylvain Levi
T’AI eu roccasion, il y a trois ans, d’attirer Tattention sur un terme
obscur du bouddhisme Sanscrit, mdlavihara, que j’avais rencontre
dans le texte du Maha-Karmavibhanga dont je donnais I’edition et
la traduction (Paris, 1932, p. 63, note). II s’agissait, dans ce passage,
du fameux Candanamala vihara erige par Purna a Surparaka. J'avais
rappele a ce propos le recit du voyage de Bhagavat a Surparaka sur
rimutation de Purna et de la consecration du vihara, tel qu'il est
rapporte dans le Divyavadana (ii, Purnavadana) ou plus exactenient
dans le Yinaya des Mida Sarvastivadin’s (Osadhivastu, chap, ii =
Dulva tibetain ii, 40 sqq.). J'avais signale quelques autres textes,
tant sanscrits que palis, ou la meme expression se retrouve, soit a
propos du meme Candanamala vihara (Kandjour, Mdo iii ; Para-
matthadlpanl (et non Apadana ; a rectifier) sur Theragatha 187-188
Paramatthajotika sur Suttanipata, Puralasutta ; soit comme une
designation generique (Mahavastu ii, 367 ; Siksasamuccaya, p. 300 ;
Divj-avad., p. 79 et p. 467 ; texte Sanscrit de Koutcha publie par
M. Liiders, Pr. Ak. Wiss. 1930, i, p. 23). Aux te.xtes sur le
Candanamala je puis ajouter maintenant le Manjusrimulakalpa
(Trivandrum Sk. Ser., p. 619, inf.) qui mentionne un bhiksu nomme
Nanda ou Nandaka, hote du Candanamala, avec une longevite de
trois cents ans ; —
bhavitd Candanatmle ’smin hhilvsur Kandako bhuvi
tasmin kalddhame j)rapte jived varsdmlutrayam.
J’avais rappele aussi I'embarras des traducteurs, a commencer
par les interpretes tibetains qui avaient hardiment substitue maid
a yndla et rendu par phren ba “ guirlande ” ; Burnouf, dans sa
traduction du Purnavadana, avait fait de meme {Introd. d I'Histoire
du Buddhisme indien, p. 258), et Bendall apres lui (Siksasamuccaya,
loc. laud.) ; Liiders s'etait contente d’un point d" interrogation.
Les textes jainas, temoins precieux et trop peu consultes, donnent
la solution de cette difiiculte. Le mot mdla y parait souvent, non
seulement dan.s les recits et les contes, mais aussi dans les prescriptions
d’ordre rituel. Jacobi I’avait rencontre dans rAyararhga (ii, 1, 7)
et dans la traduction qu'il a donnee de cet ouvrage {Sacred Books,
vol. xxii), il en a bien reconnu la signification. II est interdit au moine
ou a la nonne en tournee d'aumones, d'accepter de la nourriture
qui a ete placee en lieu eleve, “ sur un poteau, ou un pilier, ou un
620
SYLVAIN LEVI —
treteau, ou un mala, ou une plate-forme, ou une terrasse ”
(khamdhamsi vd thambhamsi vd mamcamsi vd yndlamsi vd pdsdyamsi
vd hammiyatalarhsi vd). Jacobi a rendu mdla par loft ”, le grenier
en haut de la maison. Et il met en note : “ The word is not explained
in the Tika and Dipika ; the Guzerati translation says that the word
is lokapratita, commonly understood. It is probably the Marathi
mdl or mala ; the former word denotes a loft floored with bamboos ;
the second, the room formed by overlaying with shght sticks the cross-
beams of a house, a loft, an erection or stand in a cornfield, scaffolding
(of a building). Molesworth, Marathi and English Dictionary, s.v.”
L’excellent dictionnaire pracrit de Pandit Hargovind Das T. Sheth
{Pdia Sadda Mahannavo, A Comprehensive Prakrit-Hindi Dictionary,
Calcutta, 1923) distingue deux homonymes mdla, masculins tous deux,
provinciaux {desl) d’origine tous deu.x ; a I’un il attribue les
significations de 1° jardin {drdm, bagicd) ; 2°, plate-forme {mane,
dsan-vises) ; 3° harmonieux (manju) ; pour I’autre il donne comme
significations : 1° nom de pays (des-vises) ; la litterature sanscrite
connait ce nom ; 2° partie superieure de la maison (yhar kd upari-
bhdg), support {tala), et il ajoute : “ En langue du Guzerat, e'est
mdlo (Giijardtl me mdlo). Il indique encore comme troisieme sens :
une espece d’arbre {vanaspati-vises). Et pour toutes ces significa-
tions, il donne des references precises a des ouvrages jainas.
L’admirable encyclopedic du jainisme ^vetambara, I’Abhidhana
Rajendra une oeuvre modele, qui devrait servir d'exemple au
bouddhisme et au brahmanisme, et qui n’a pas encore obtenu la con-
secration qu'elle merite enregistre le mot mdla, masculin, avec les
interpretations suivantes (en Sanscrit) : 1° la partie d’en haut
(uparitanabhdye) ; 2°, plate-forme (imncddika) ; 3° nom d'un pays
montagneux. Et il rappelle les significations indiquees par Hema-
candra dans sa Desinamamala : jardin ; harmonieux ; plate-forme
{drdmanutnjumancesu). Les deux lexicographes, a la suite de
Hemacandra, qui suit les commentateurs jainas, se servent, pour
expliquer le mot mdla, de ce mot maiica que nous avons \’ti paraitre
a cote de lui dans le passage cite de TAyaramga. A la difference de
mdla, manca est un mot d'usage courant dans la litterature sanscrite ;
la formule manedh krosanti est un des exemples stereot\q)es de
metonymie qui reparait a I'infini dans tout rAlarhkaraAastra. P. K.
Acharya, dans son utile Dictionary of Hindu Architecture (oil il n'a
pas malheureusement mis a contribution les ouvrages jainas, qui out
tant a fournir dans ce domaine) explique manca par “ bedstead.
MALA VIHAKA
621
couch, bed, sofa, chair, throne, platform, pulpit ” et cette multiplicite
d’equivalents atteste I’extreme elasticite du sens. L’Abhidhana
Rajendra dit fort bien, s.v. : “ un manca est fait d’un lit de bambou
pose sur des montants (sthunandm upari sthapitavarhsakatakddimayo
mancah) ” et il ajoute : “ Tout le monde sait ce que c'est {loka-
prasiddhah).” Et il a recours encore, pour I'expliquer, a un derive
du mot mala, nidlaka : “ c’est un mdlaka oit des gens qui veulent voir
un spectacle peuvent s'asseoir (preksanakadrastrjanopavemnanimilte
rndlake).” Et, sous le mot mamcdtimamca, il repete : “ manca est un
mdlaka ou des gens peuvent s'asseoir pour regarder une grande fete
{mahotsavavilokanajandndm 'upave.sananimittamllako mancah)’’ Le
commentateiu du Sthananga ajoute encore une precision importante ;
le texte (3, 1, 145; p. 124a) mentionne toute espece de grains qui
sont conserves dans des endroits divers, entre autres “ conserves
dans un manica, mamcdutia [dgupta], conserves dans un mdla
(rndldtitta) Le commentaire explique le premier mot dans les termes
que I’Abhidhana Eajendra lui a empruntes pour sa definition :
sthundndm upari sthdpitavarhsakatakddimaijo jnnapratltah ; pour le
mot mdla, il en est de meme : mdlaka grhasyoparitanahhdgah ; mais
il ajoute une reference a un texte jaina, dictionnaire ou commentaire ;
“ Et il est dit ; un tnamca n'a pas de murs, et do plus un mdla est
au-dessus de la maison (akkuddo hoi mamco mdla ya gharovarim hoi
glose en sancrit par : akudyo bhavati mahco mdlas ca grhopari
bhavati.” Ainsi le mdla. comme le manca, n'a pas de mur ; mais a
la difference du manca, il est situe en haut do la maison ; on con 9 oit
des lors que les deux mots servent a s'expliquer mutuellement ; un
mdla est un mirador, une loge, un kiosque, un pavilion, un balcon
place en haut de la maison, et en general place en haut. C’est dans
ce sens que ce mot — sous sa forme derivee : mdlaka — parait dans un
texte Sanscrit jaina, la Simhasanadvatrirhsika editee par Weber,
Ind. Stud., XV, p. 266 sq. ; Dans un village d’Avanti, un brahmane
tirait de son champ des moissons merveilleuses. '• Alors il fit elever
dans le haut de ce champ un mdlaka. Or chaque fois qu’il monte dans
ee mdlaka, il se sent de la grandeur ; chaque fois qu'il en descend, il
se sent miserable.’’ Il va a la capitale informer le roi Bhoja de ce
prodige. “ Le roi vint lui-meme examiner les lieu.x. mais il ne remarqua
rien. Alors il monta en personne dans le tndlaka et il se sentit une
grandeur extraordinaire. . . . Le roi acheta a prix eleve ce champ au
brahmane, et il fit creuser au-dessous du mdlaka.” C'est la qu’on
decou\ue le trone merveilleux aux trente-deux statues autour duquel
622
MALA VIHARA
se deroulent les contes. II est bien clair qu’il s’agit ici d'uu de ces
abris portes sur dcs poteaux qu’on voit si souvent dans la campagne
indienne, on le paysan se repose a I’ombre d’un toit de chaume et
surveille ses cultures. Weber avait bien devine (ib., p. 219) qu’il
s'agissait “ probablement d’tine estrade ” mais qu’il imaginait ronde
(wobl einena ringformigen Gestell) sans doute sous I’influence, Ini
aussi, du mot feminin mold “ guirlande
Examinons maintenant la miniature nepalaise pubbee par
M. Foucher dans son iStude sur V Iconographie bouddhique de I’Inde,
pi. I, 6 ; j'ai deja indique dans ma note du Maha-Karmavibhanga
qu’il fallait rectifier la transcription et lire, au lieu de Supaca, Supara :
Supdranagare VuntraJca (M. Foucher a lu ; Vulbhuka) vltardgakrta
Candanavihdra. Supara est une de formes que prend le nom de la
ville oil Purna eleva son fameux Candanamala vihara ; c'est sous
cette forme meme Zovnirdpa que Ptolemee, vii, 1, 6, ecrit le nom
du grand port que frequentait le commerce d’Alexandrie. J’ai
anterieurement propose de considerer I'enigmatique Vuntraka (1)
comme une alteration graphique assez facile a expliquer de Purnnaka.
Quoi qu’il en doive etre de ce nom mysterieux, le monastere de
santal a Supara ne pent etre, dans la galerie des lieux saints les
plus celebre.s, que le Candanamala vihara. Et la miniature montre
en effet au premier plan une loggia, une espece de balcon convert
accroche au flanc d'une maison d'habitation ; dans la loggia se tient
un personnage portant une guirlande au cou, le meme peut-etre
qu'on voit apparaitre au fond, depassant largement de sa taille
demesuree le stupa a parasol et le pilier surmonte d’une image (de
lion ?, pour rappeler le Sakyasimha, le Lion des Sakya’s ?) un ensemble
qui evogue necessairement le souvenir des grands vihara's du Nepal.
Est-ce le Bouddha qui vient sur I’initiative du Purna et qui ensuite
visite le monastere ? L’artiste le fait alors paraitre dans la loggia de
santal qui avait valu au monastere sa reputation. Nous avons done
la I'illustration d'un mdla. Et il est frappant de voir im mot considere
comme proprement gujarati paraitre des les origines du bouddhisme
dans le nom d'un monastere qui etait la gloire du Guzerate. II y aurait
la un indice curieux de I'anciennete du vocabulaire des parlers
regionaux de I’lnde.
Je suis heureux de soumettre cette question au grand savant et au
vieil ami dont le nom reste et restera toujours lie indissoblument a
I’etude linguistique de I’Inde, au createur et a I’auteur du Linguistic
Survey of India, Sir George Grierson.
Nochmals mieccha
By B. Liebich
TTEBER diese Entsprechung des griechischen bdrbaros im Sanskrit
^ habe ich im 72. Band der ZDMG. (1918) Seite 286/7 eine kleine
Studie veroffentlicht. Den Anlass dazu gab ein Aitikel von K. P.
Jayaswal, der am gleichen Ort einige Jahre vorher erschienen war.
Jayaswal hatte richtig erkannt, dass die allgemeinere Bedeutung
von mieccha, wie sie in der Literatur erscbeint und auch von den
indischen Grammatikern angegeben wird, sekundar sei, und dass das
Wort urspriinglich ein bestimmtes nichtindiscbes Einzelvolk bezeichnet
haben miisse : ,, like Yavana,“ sagt er, „ Mlechchha is a foreign word,
and like Yavana it originally meant a specific foreign people."
Das griechische bdrbaros, lat. barbarus bietet in dieser Hinsiclit
kein Problem ; es zeigt in seiner reduplizierten Form deutlich seine
Abkunft von einer malten lautmalenden (onomatopoetischen) indo-
gerraanischen Wurzel bar, die sich, ebenfalls redupliziert, als Verbum
mit der Bedeutung plappern, summen u.a. in mehreren slawischen
Spracben und im Litauischen erhalten hat. Ja diese Wurzel bar hat
in der neudeutschen Theatersprache eine interessante Wieder-
auferstehung erlebt, wo sie in der Vermummung Barbara oder
Rhabarber vom Sprechchor allgemein als ,, Volksgemurmel “ Ver-
wendung findet. Neben indogermanisch bar steht die Wurzelvariante
bal in lat. balbus ,, stamnrelnd lat. balbutire und skr. balbald-karoti
„ stammeln, stottern also in Beziehung auf einen individuellen
Sprachfehler. Griechisch barbaros bezeichnet dem gegeniiber von
Haus aus einen Menschen, der sich durch seine undeutliche oder
unverstandliche Sprache als Volksfremden zu erkennen gibt.
Soweit kaim man also Jayaswal zustinmien, aber seine Verknupfrmg
von mieccha mit einem bestimmten Volk ist allzu kiihn und durfte
nicht unwidersprochen bleiben. Er geht namlich vom altesten
Vorkonimen des Wortes im Satapatha-Brahmapa aus, wo gesagt wird,
dass die Asura mit dem mleccha-Knl helavo kelavo im Kampf unterlagen
und zu Grunde gingen. Die Asura sind ihm (wie anderen neueren
Hindu-Gelehrten) die Assyrer, in helavo findet er das hebraische
ehah ,. Gott “ und in mieccha selb.st das semitische melech ,, Koni^ “ '
Was die Asura betrifft, so geniigt es wohl, daran zu erinnern. dass
mit diesem Wort in den alteren Buchern des Rgveda die eigenen Gotter
624
B. LIEBICH —
(India, Varuna, die Aditya’s) bezeichnet werden ; erst vom zehnten
Buch ab, im Atbarvaveda und in den Brabmana's sind die Asura zu
Damonen geworden und steben den Deva’s als Feinde gegeniiber.
Die umgekebrte Entwicklung haben diese Worte bekanntbcb im
Iranischen genommen, wo asura, lautgesetzlich als ahura erscbeinend,
im Namen des hochsten Gottes Ahuramazda — Ormazd wiederkebrt,
wahrend bier die Diven zu Damonen und Gotterfeinden geworden
sind. Martin Haug, der zuerst auf diesen Gegensatz bingewiesen hat,
sab darin das Anzeichen einer religiosen Spaltung zwischen Indern
und Iraniern, weicbe die Trennung der alten Indoiranier in zwei
gesonderte Volker begleitet oder veranlasst babe, was wohl auch jetzt
noch die natiirbcbste Deutung bleibt. Jedenfalls bat Asura nur mit
den arischen Persern, nicht mit den semitischen Assyrern etwas zu
tun, und damit erledigt sich die Gleichung mleccha mit melech von selbst.
In meiner oben erwabnten Entgegnung unterstricb icb die Tatsache,
dass das spater so haufige Wort mleccha in der Tat in Eg- und
Atharvasamhita wie auch in der vedischen Prosa des Yajurveda
und der alten und mittleren Brabmana’s vollig feblt und erst im dritten
Bucb des (verhaltnismassig) jungen ^atapatha-Brabmana erscbeint.
Im Rgveda bis zura zehnten Bucb einschliesslicb finden wir sogar eine
andere Entsprechung von bdrbaros, die spater ganz verschwindet,
namlich mrdhravdc ,, dessen Rede (Sprache) mangelhaft, — fehlerhaft
ist, kauderwelscb — , barbarisch redend “ (Geldner, Der Rgveda in
Ausivahl, i, 138). Als Redaktor des Satapatba-Brahmana wird uns
Yajnavalkya genannt, der am Hole des Konigs Janaka in Videba,
dem spateren Magadha, im unteren Gangeslande seine beruhmten,
im Brahmana selbst aufgezeicbneten Disputationen uber Atman
und Brahman bielt ; und im ersten Buch desselben Brahmana (i, 4, 1,
10-18) finden wir die Erzahlung von Videgha Mathava, dem
sagenhaften Stammvater der Videha’s, und seiner einstigen Besiedlung
dieses Landes, das ganz felderlos (aksetrataram) und sebr fliessend,
sumpfig {srdvitaram) gewe.sen sei, bis Agni Yaisvanara, gefolgt von
Yidegba Mathava, es iiberbrannt babe (atidaddha). Bei Agni
Yaisvanara, dem ,, Jedermannsfeuer wird man an das Abbrennen
des Jungle zu denken haben, das der Besiedlung und Bebauung
des Landes vorangebn mu.sste. Hier am unteren Ganges diirften wir
also nach alien iiusseren Indizien den Ursprung des Xamens Mleccha
zu suchen haben.
Das Wort ndeccha im Sinne des griechischen bdrbaros ist nicht nur
baufig im klassischen Sanskrit, sondern ist auch in der gleichen
KOCHMALS MLECCHA
625
Bedeutung in die indischen Dialekte iibergegangen. Konsonanten-
gruppen werden tier durch Assimilation oder Vokaleinschub beseitigt ;
der durch Assimilation entstandene Doppelkonsonant wird im Anlaut
durch den einfachen ersetzt. Dazu kommen hier und da, namentlich
wo es sich um die Wiedergabe von Fremdworten handelt, noch
sporadische Veranderungen. Demgemass finden wir fiir mlcccha im
Pali milakkJia, in SaurasenI, Jaina-Maharastrl und Apabhraihsa
meccha, in ArdhamagadhI sogar vier Formen : miliccha, milakkhu,
meccha und miccha nebeneinander. Alle diese Formen sind durch
Parallelfalle lautgesetzlich vertretbar ; die Nachweise sind bei Kuhn
und Pischel leicht zu finden.
Auf Grimd dieses Tatbestandes vvies ich auf die ilech hin, ein
nichtarisches Volk von jetzt rund hunderttausend iMenschen, das heut
hauptsachlich im Taral, also im Jungle, westlich des Brahmaputra
haust, teds in Assam, teils in Bengalen. Obwohl sie durch Hautfarbe
und mongolide Ziige den Bodo nahe stehen, haben sie keine Tradition,
dass sie jemals ausserhalb des Jungle gelebt haben. Ich erinnerto
schliesslich daran, wie stabil haufig die Namen auch der nichtarischen
Volker in Indien sind ; die Bliilla und Kirata der alten Sanskrit-
Autoren finden wir noch heut in den Bhil des Vindhya und den Kiranti
des Himalaya wieder, die Sahara des Plinius und Ptolemaeus unter
ganz dem gleichen Namen in den ostlichen Ghats, die Darada des
Mahabharata in den Darden von Dardistan.
Als ich meine Studie veroffentlichte, kannte ich die Jlech nur
aus dem vortreffhchen Werk von Sir Athelstane Baines, Ethnography
(castes and tribes), Strassburg 1912, p. 129, von dem, beilaufig bemerkt,
mein Kollege Freiherr von Eickstedt erst kiirzlich bedauerte, dass es,
weil im Grundriss der indoarischen Philologie erschienen, bei seinen
Fachgenossen, den Ethnologen und .knthropologen so gut wie
unbekannt geblieben sei. Ich war aber hocherfreut, als ich einige
Jahre spiiter in der Linguistic Survey of India, dem monumentum
acre perennius des ehrwiirdigen Altmeisters der Indologie, zu dessen
Ehren diese Festschrift entstanden ist, in vol. iii, 2, p. 1 die folgende
Stelle fand : ,, The Bara folk who live to the west of the Kamrup
district are called Mech by their Hindu neighbours. This word is
probably a corruption of the Sanscrit ‘ MIechchha which corresponds
to the original meaning of our word ‘ Welsh ", i.e. foreigner, stranger.’'
Es kann fiir uns dahingestellt bleiben, wer von beiden den Tatbestand
zutreffender erfasst hat, die Linguistic Survey des Census von 1901
oder die Ethnographical Survey des gleichen Censii.s. auf deren Angaben
626
NOCHMALS MLECCBA
Baines sich stiitzt. Jedenfalls erhalt meine damals noch ziemlich
schiichtern geausserte Vermutung durch diese Stelle eine wiclitige
Bestatigung.
Noch ist eine dritte Behandlung dieser Frage zu erwahnen, die
ein Jahr n a c h meiner Arbeit von Isidor Scheftelowitz am
gleichen Ort {ZDMG. 1919, S. 243/4) verdffentlicht wurde. Sch.
verwendet eine andre Methode und kommt zu einem andern
Ergebnis. Er trennt pali milakkha ganz von skr. mleccha und weist
sie zwei verschiedenen, aber indogermanischen Wortsippen zu.
Allerdings geht es dabei nicht ohne Gewalttatigkeit ab. Zunachst
ist es ziemlich kiihn, die Zusanimengehorigkeit dieser beiden Worte
zu bestreiten, die genau im gleichen Sinne in der Literatur verwendet
werden imd durch die oben erwahnten Zwischenformen verbunden
sind, weshalb alle griindlichen Kenner der indischen Dialekte, Kuhn,
Franke, Stede, Pischel, wie auch die einheimischen Grammatiker an
der Zusammengehbrigkeit festhalten. Die Tatsache des spaten
Auftretens von mleccha, die mir wichtig erschien, wird von Sch. mit
Stillschweigen iibergangen.
Skr. mleccha muss sich zunachst eine Umformung gefallen lassen.
Sein palataler Ausgang soil auf einen alteren dentalen zuriickgehen.
Durch diesen (sehr ungewohnlichen) Lautwandel kommen wir
auf ein indogernianisches *mlais-sko, das auch in cymrisch bloesg
vorliegen und mit lat. blaesus aus *mlais-sos verwandt sein soil. Diese
beiden Worte bedeuten ,, stammelnd, stotternd “. Mlais-sko ist aber
eine rein papierne, kaum aussprechbarc Konstruktion.
Pali milakkha wird von Sch. mit skr. mftrkha ,,Dummkopf“
vereinigt und zu einer indogermanischen Wortsippe gestellt, deren
Grundbedeutung ,, erstarren zu sein scheint, wahrend die ihr
zugeschriebenen Yerben in den verschiedenen Einzelsprachen meist
schweigen oder schlafen bedeuten.
Scheftelowitz schliesst mit den Worten : Die Untersuchung hat
somit ergeben, dass altind. mleccha nicht mit altind. *mlska, *mlsku
(pali rnilakkho, prakr. milakkhu) etvmologisch verwandt ist, aber
beide Worte rein indogermani.schen Ursprungs sind. — Ob seine
Ausfiihrungen iiberzeugend genug wirken, urn ein so zuversichtliches
Urteil zu rechtfertigen, mogen andere entscheiden.
Nugae Burushaskicae
By D. L. E. Lorimer
T he makers of Burushaski, whoever they were and wherever and
whenever they lived, were people of resource and ingenuity,
perhaps of imagination. So much is evident from the grammar of the
language, which I have described in some detail elsewhere. The
Burushaski Language, vol. i, Oslo, 1935. I am not referring to the
modern speakers of Burushaski. They appear to me lacking in these
quahties. They might be gifted as craftsmen, but not as creative
artists.
After further researches I would now also attribute to the begetters
of the language some subtility of mind. I would even credit them with
a measure of originality ; but if I adduced in evidence the phenomena
which I am here going to describe I fear I should be speedily convicted
of ignorance of all the most interesting languages of the globe.
1. I will merely mention here in passing the system under which
by varying the quality, stress, and length of the vowel of its pronominal
prefix (or infix), a verb may be changed from Intransitive to Simple
Transitive, then to a Transitive where a third party is affected (to
whom the pronoun-prefix then refers), and lastly into a Causative
Active. The Intransitive form of the verb sometimes has no pronoun
prefix. In the following examples the pronoun-infix is in the 3rd
person singidar form, urdess otherwise stated : — ■
Intrs. soqa dis'lLimi the cloak became wet
Trs. 1. ine ja soqa d'esilimi he wet my cloak
Trs. 2. ine ja soqa d'a silimi he wet my cloak for me
(1st pers. sg. pronoun-infix)
Cs. u’ge gu'imo nokor ja soqa you made your servant wet my
d'e asiloma cloak ; or
u'qe gu imo nokor ja soqa you made your servant wet my
d'a’Asiloma cloak (for me)
(1st sg. pronoun-infix.)
In the Transitive 2 and the Causative the vowel is markedly long
and may be diphthongized. There is probably also a change in tone
between the vowels of Trs. 1 and Trs. 2 and Cs. Neither I nor my
informants could decide.
2. I will again only briefly refer to a number of Intransitive
628
D. L. K. LORIMEK
verbs wbicli add, or may add, a theme (apparently -ia-) to the root
when the subject is plural : —
du’SAs Pres. Base du's- with sg. or pi. subj.
duwASE.As Pres. Base duwASAc- with pi. subj. only,
both meaning to come out, emerge, etc.
inE du'saei he is coming out
u’E du’sa n or duwASAca’n they are coming out
tASwi'r su’a d'usila the 'photograph has come out well
tASwi’riq su'a d'usitsa or duwAsa hitsa the photos have come out well
Other such verbs are : —
gaTtSAS
to run
gartsimi
he ran
(gartsomAn
they ran)
garc'aei.AS
garcAmAu
they ran
horu’tAs
to sit
hurutimi
he sat
(horutomAn
they sat)
horuc'aei.As
} >
horucAiDAn
they sat
the two preceding verbs the form with the -^a-
■ theme is usually
only employed in the past tenses (plural), where it ordinarily displaces
the simple form.
guiyAS II giyaei.AS
*-uyAs with h ot X subject |
buyAs with y subject )
*-uyaei.As with h ox x subject pi. j
buyaei.AS with y subject pi. j
to fall, plunge, etc.
to become dry
to become dry
h, X, and y refer to different categories or “ genders ” of nouns. An
indicates that the verb is preceded by a pronoun-prefix.
asterisk
I can quote one Transitive verb, more will probably be discovered,
where a -ia- theme appears when the object is y plural : —
pusa s I object
pos'uyAsj ® ■’
*-pfusAS with h ox X obj. sg. or pi.
pus'aei.AS with y pi. obj.
All meaning to tie up.
Thus : —
bAlda pos'e, or pos'u
hAyur ipfos
korsimots upfos
bAldAfl (// pi.) pos'a
siqa (y pi.) pos'a
tie up the load
tie up the horse
tie up the chairs
tie up the loads
tie up the grass
NUGAE BURUSHASKICAE
629
3. I now pass to phenomena which require less technical knowledge
to make them fully intelligible.
Burushaski has a system of depreciatory or contemptuous
expression which in certain circumstances is apphed to many of the
commonest objects of daily life.
The depreciatory effect is produced by attaching a qualificatory
word to the ordinary word denoting the object. Where the meaning is
unmistakable the qualificatory word is sometimes substituted for the
standard word. This qualificatory word is sometimes the name —
(a) of a particular part of, or of a small quantity of the object ;
(b) of a small or inferior form of it ;
(c) of a small or inferior object made of, or associated with it.
Sometimes it is merely —
(d) a more or less general word denoting smallness in quantity
or size.
The following are a few out of a large number of available examples,
falling roughly under the four above headings. (The suffix -An, -n,
denotes “ one ”, “ a ”).
Simple word.
(а) hu'co soft leather
hoot
SAP horse-shoe
yAs'il (Jire)u’ood
zu warp yarn
y'e iq grapes
(б) hAy'ur horse
ts.hi’r she-goat
ha house
bir'Anc mulberries
(c) gAp hide
pf'Ata bold
p'Aci cotton cloth
bAt shin, leather
Depreciatory term.
huco iltumAl
sApE mu s-An
yAsi le joco r-An
zu.a yay-An
ye iq£ chu-An
ye'iqe pfulpful-An
hAy'ur bodo ku-An
tS.hi TE duw-Au
ha duku ri
ha goti
biTAnc mAyaro ti
gApE Ask'i ltar
pfAta cok'uli
pAci.E gAk’a y,
pAci.E tAt'Ay
bAtE chAya (lukan)
bAtE barp'i t (lukan)
SAPE g'ili-An
Meaning of
qualificatory word,
lit. “ ear ”, secondarily
“ side piece of boot ”
an end, an edge
a shaving
one long thread of yarn
a bunch
a berry
a small pony
a hid
shelter, hut
hut
unripe mulberries
thong for fastening yohe
to shaft of plough
fragment of bowl
patch, old scrap
ditto slightly larger
patch
strap
a nail, peg
SAP horse-shoe
There are not many examples of this type.
630
D. L. R. LORIMER
Meaning of
Simple word. Depreciatory term. qualificatory word.
(d) sAp'ik bread sAp'ik pfu'k-An a tiny piece
sAp'ik lAp-An a mouthful
SAp'ik coru’k-Aii a hroken-off piece
ts.hil water ts.hil thi's-An a little {liquid)
del oil del yAsa'-n a little
dsl 3 ni'w-An a tear
With words denoting cloth, or articles of cloth, the terms loq scrap,
bod'ol rag, are commonly used.
These depreciatory expressions have different values according to
the context in which they are used : —
1. When used of a personal possession they minimize the value
of the article and indicate humility, or the absence of self-magnification,
on the part of the speaker. So one may say : —
ye’iqe pfulpfulAn go r ditsa ba I have brought you some grapes
{“ a grape berry ”)
bu’yomots adta mAni.En There have become two little
bunches, i.e. I have got a couple
of stacks (a lta huTsseints) of
cut crops ready for threshing
sarmotsAn pfAlo sedas aeip its.haei They reckon it reprehensible to
ba’n ; kuto.An pfAlo seiban say “ (I have) a sack of grain ,
they say “ {I have) a small bag
of grain ”
2. They may be used of another person’s things with the effect
of insulting or depreciating him ; —
b'ESE gosE dukuri.olo horu’ta ? Why have you taken up your
abode in this hovel ?
ja ha kE u'qE dokuTi hAna ? Are my house and your hovel one
and the same thing ?
They are thus used when expressing dissatisfaetion with a gift
and the giver : —
Q. thAmE hAld'EUAn guwaeya? Has the king given you a {he-)
goat ?
A. bESAnE hAldEn ? bi'ske {Yes but) what sort of a goat? He
Dsko ayonsei has given me three hairs (i.e.
a goat in very poor condition)
birAncE mainAU ach'i baei He has given me a “ stump ” of a
mulberry tree
NUGAE BURUSHASKICAE
631
3. These uses are straightforward ; the subtility to which I have
referred appears iu the next, in which by depreciating the object
the owner or giver is exalted ; the implication being that while, e.g.
the coat is a perfectly good coat and to the recipient a glory, to the
giver it is a mere scrap of cloth or a tattered old garment. Or, the
roles being reversed, the full meal offered to the guest is to him merely
a mouthful of bread or a cup of tea. Again, out of courtesy to the
recipient the giver will represent his gift as small and unworthy.
Thus I say to a superior : —
hu’co idtumAl Achi. Give 'me a side of a boot, i.e. give me a pair of boots.
and the giver will say as he gives them ; —
i'ltumAliq bitsAq. gotse yAski k£ ap'i.EU. The'y are {a) mere {set of)
bootsides. They are not tvorthy of you.
guwAtAtom menAU jo. Give me a set of old clothes off your person.
ko te bud'ol lokAU Achi. gAka yAu bila, da u qar besAn ? Give me
a rag of a coat. It's a mere old scrap, and so xvhat is it to you ?
'aya Giltar aic'Am “ hu cotse WAscar gApe pfo lco lukAn bi ke ”, nosen
daTsei. Ju ju, lukAn ti’ko miy'uwin. My father icas going to
Gilgit, and saying “ if there is a little scrap of hide to apply to my
boots” , he has sent me {to you). Many salams ! Give us a hit.
ja gAp guy Am seiba. besAUE gap ? pisok lukAU bilum. AkhurumAU
Aski'ltar bilum. You say I gave you hide. What sort of hide {was
it) ? It was only a little calf-slin. A little bit of thong like this.
bAtiq thuTAu aT thi l 'tti.a. Will you he so good as to sprinkle out
{■with the fingers) a little dusting-flour for me ? i.e. will you give n\e
some flour ?
Here “bAtiq”, “ thuTAU ” and “thil” are all in their different
ways diminutives.
This conception of depreciation is carried over into the sphere of
the verbs. Thus one says : —
huco ulta. Put on the boots (u-, 3rd pi. pn. pf. agreeing with hu’co).
This is normal and grammatically correct, but may be regarded
as honorific to the boots. Or one may say ; —
hu'co ilta (i-, 3rd sg. pn. pf.).
This is humiliating to the boots, but honorific to the person
addressed. Doing honour to another person and humiliating the boots
one says : — •
ggjsE bu't cu:k Eti. Put on this hoot
632
D. L. R. LORIMER
To a person of distinction a would-be host says
ja hadar chAm mAHE ! gOT cad o’CAm. Enter my house (i.e. Do me
the honour of entering my house). I'll rruike them make tea for you
(i.e. ril entertain you to a meal).
On the other hand, however : —
ya ski sisAn 'e ts.huyAsar xe.a’lAte uxAtAtom guts seiba n : “ mi
hadar chAm menma ? ” With the intention of not taking an
unworthy jperson (into one's house) they say from their lips (lit.
mouth) : “ Will you do me the honour of entering our house ? ”
4. The last phenomenon to which I will here call attention is
the existence of an indefinite number of verbal compounds (an
adjective or noun plus an au.xiliary verb) of which the meaning is
modified by alteration in the vowel of the first component. The
variants from the standard are in the nature of diminutives.
As a typical example the following may be given ; —
Normal. Variant.
SAT etAS to roll out (dough) |iT etAs to roll out a small quantity
thin, with quick short motion.
saT EtAs to roll out larger (?) quantity
slowly.
This represents, I think, the commonest series of vowels, and in
it the i’ vowel seems to point to a diminishment in the action of the
verb with regard either to the thing acted on or the means employed,
or in what results from the action. The a' vowel may also denote some
quantitative reduction, but seems frequently to indicate slow motion.
There seemed to me to be a difference in the tone of the vowels,
the tone of the norm being normal, that of the i’ high, that of the
a’ low.i The effect of the variations may be intensified by further
lengthening those vowels. Many of the first components may also be
reduplicated, giving probably the sense of repetition or continuance
of the action.
In some cases, where the vowel of the norm is -a-, there are
diminutives in o’ or u'.
When the vowel of the norm is i or u the diminutives are obtained
by simply lengthening and lowering the tone of those vowels.
Frequently where the norm has -i- there is a diminutive also with -a-,
and where the norm has -u- there is a diminutive with -i-.
^ In the following examples a line above a vowel denotes a high tone, and a line
below a vowel a low tone.
NUGAE BUEUSHASKICAE
633
It will be evident that it is impossible from the vowel of any
isolated form to say whether it is a normal or a diminutive.
The following are a few miscellaneous examples. I cannot in all
cases given the precise meaning of the diminutives, either vis-a-vis
the norm or vis-a-vis each other : —
thAm thAm EtAs
thi m thi m e.)
to sweep up a small
to sweep up
thithi m e. j
quantity
tha m £.
to sweep up a small
quantity, or slowly
XAlAt EtAS )
ZAlaXAlAt E.j
to work round a little
xi lixili’t E.
with one finger
to work round with the
XATa t E.
ditto, a few times.
hands dough or por-
ridge in a vessel
slowly.
dAq EtAS
di q EtAs
to make hard, to cook
da q DOLAni bi
it has become a little
hardened, slightly
cooked.
Ias EtAsl ^ ,
„ ito lick
li'S EtAS
IaU? e. /
lil'fs E.
la'I E.
bull lUAmu se'AstsE at'ayseya ba ; baIaI
la s EtAstsE da yseya ba.
I’m not worried at the cat’s drinking the milk. I’m annoyed at its
{way of) slowly licking its moustache.
This may be said when a new man receives an official appointment :
“ I am not concerned about Smith’s taking bribes. I can’t stand his
putting on airs.”
(There is uncertainty regarding the exact meaning of da'yseya ba.
It was explained as meaning “ I am afraid of ”. I have not, I think,
met it elsewhere. It is probably related to a’yseya ba, “ I regard as,”
“ consider.”)
sAq etAs siq e. )
to ccmd> sis'iq e.
\to co7nh ivith short quick tnovetnents
sa:q e. 1
i V [ditto, with long sloiv movements
sasa q e. J
Examples of other vowel patterns are : —
lu k EtAS li k E.
to jmt a mouthful into Imk e.
one’s mouth
VOL. Vin. PARTS 2 AXD 3.
41
634
D. L. E. LORIMKR —
thi l EtAS thAl £.
to Sprinkle or flick, thi:l £.
flour, sugar, etc., tha:l £.
with the fingers
suk EtAs) si’k £. I
susuk £. j ^ sisi'k E.J
sa:k £. I
susuik E.J
thar EtAS I to undo, thir e.
thAthar e. f untie thor e.
tho:r £., thothoT e.
tha;r z.
tho:r and tha:r are polite words :
hayuTE gAsk tho:r e. Would you be so good as to take the trouble to undo
the horse's rope
tha:r Efiam. I'll undo it with pleasure ; it's no trouble
la s EtAS los £. I
to stnear lol'o s e. )
plaster over la:s £. J
lulu s £. I
ga l lu:s maniuii. The wound re-opened {sloughed ?)
This vowel-variation is applied in other cases besides the verbal
compounds which have just been considered : —
ts.hilE thi san
a little ivatcr
thiisan |
th'Asan - a very little
th'a'san)
lukan a little
kAmAn a little
sag manimi
it became light, day
dawned
yaski worthy
ImkanJ
li kan J
a very little, rather little
ka man ,, ,,
sig manimi \
ssvg „ I peep of dawn {?)
si:q „ j
sisa/q „ '
ya'.ski (so little worthy as to be) unworthy
Lengthening of the vowel in ordinary verbal forms introduces an
element of doubt or uncertainty ; —
Akhi s'Enimi he said so (in the ordinary, positive way)
NUGAE BURUSHASKICAE
635
Akhi s'E'nimi he said so (wieniphatically, not clearly), or he said smne-
thiny to that effect, or the statement he made was doubtfully true.
SEna b'seyAin I had said. . . .
s's’na bseyAm perhaps I had said (something like that)
bitAn mowAsca-n seiba'n (people) say they are going to make the
“ Utan ■’ dance
bitAn mowAsca’n s'e’iba’n. be'yAin (people) say they are . . ., etc.
But I don’t know.
Akhi Et'Am he had done so
Akhi 'e;tAm he had perhaps done so
In adjectives and nouns it reduces the original significance of
the word : —
son
Y'Ayu
yot
hum'Alkum
YAl'is
jAtAn baei
gAt
blind
so:n
lame
YAy'u:
deaf
Yu:t
quickly
hum'alkum
ill
YaIes
he is old
ja tAH baei
(knot), en-
ja ka g^t 1
mity, grudge
somewhat blind
lamish
deafish
a little quickly
slightly indisposed
he is oldish
he has a slight enmity,
grudge, against me.
The use of these “ jotig borig ", “ little words," received or
improvized. may be applied to various practical ends. The following
“ true story ” is related : —
Fifty Levies went to Chitral with Wazir Humayun Big. The men,
by the time they had arrived at Drasan, had become very hungry.
They poured into the houses and gobbled up everything that came to
hand. The people of Drasan came to the Wazir and complained with
tears : " Your Levies have emptied out all the flour in our houses and
have left us nothing."
The Wazir was very angry and assembled the Levies and told
them off. Then Khujunoy Deru. speaking all in “ small words ”,
made the Wazir laugh and turned away his wrath. He spoke on this
wise : "0 Wazir. mav we be your sacrifice ! We went into the house
of an old woman, and on looking we saw that there was a very little
flour in a corner (of a sack or bin). Scraping it together we worked it
round into dough and popped it into our mouths and so saved our
lives.”
636
NUGAE BUBUSHASKICAE
This in ordinary language would run : —
Wa WAzi rE n'Azar I Mi but ch'AminE nimi mAU hin jAt gos'Anmo
ha lar nim'E n bar'e yEn ke y'u ki.Anolo th'urAU dAy'o Aq bitsum. I kE
tbAin thAm ns, x'AlaxAl'At n£, dA^ ns, lu k he, ji d'EspAsumAU.
What he actually said, no doubt in a small and pathetic voice, was :
“ Wa WiziTE uAZir ! Mi but ch'iminE nimi mAn, hin jat gos'Anmo
ha lar nimi n, biri An kE y'iiki-Anolo th'i rAn b'i tio bitsom. I kE thi m
thi m HE, xi lixili t he, di g nE, li k he, ji d'EspAsomAn.”
Zur Schrift und Sprache der Kharosthi-Dokumente
Von Heinrich Luders
TVIE Ausgabe der von Sir Aurel Stein in Ostturkestan gefundenen
Dokumente in Kharosthi-Schrift ist eine Leistung, die jedem,
der sick mit ihr beschaftigt, ehrliche Bewunderung abnotigen muss.
Allerdings wird es nocb langer Arbeit bediirfen, ehe diese sprachbch
wie sachlich gleich wichtigen Urkunden dem vollen Verstandnis
erscblossen sind. Vielleicht werden auch die folgenden Bemerkungen
als ein kleiner Beitrag zu diesem Ende von Interesse sein.
Fiir IjM, bzw. Ijd, werden in der Schrifttafel XIV unter Nr. 219-221
drei Zeicben gegeben. Von dem ersten wird S. 316 gesagt, dass es
zweifelhaft ware, ob es sich irgendwo in den Dokumenten fande,
wenn auch Ipi ahnlich geschrieben in [kd]l'pitainti in 141 vorkomme.
Da keine Reproduktion von 141 vorUegt, vermag ich iiber die Form
des Buchstabens nichts zu sagen ; Zweifel an der Lesimg scheinen
mir aber nicht ungerechtfertigt. Eine Form von der Wurzel kip
kommt sonst nirgends in den Dokumenten vor, und das Tafelchen
ist offenbar so stark beschadigt, dass sich der Inhalt der ersten Zeile,
in der angebhch [ka^pitamti steht, nicht feststellen lasst. Dies Zeichen
muss daher unberlicksichtigt bleiben.
Das zweite Zeichen erscheint nach den Herausgebern ohne den
i-Strich in alpa 468, silpa 355, mit dem i-Strich in silpiga 217.
Ausdriicklich wird auch im Index bemerkt, dass dieses Zeichen in
415 in dem Namen Ipanga verwendet ist. Nach dem transkribierten
Texte kommt es ausserdem in alpa 634, 764, jaJpiti 255 (?), 399,
jalpidemi lOQ, jalpita 524, 566, 594, jalpitamti 413, 351, jalpidavya
358 vor.
Viel haufiger ist das dritte Zeichen, das in der Ausgabe mit Ipa
umschrieben ist.^ Die Herausgeber haben S. 318 die Wahl dieser
Umschrift naher begriindet. Sie gehen von dem Zeichen 244 aus, das
in der Ausgabe durch spa wiedergegeben ist. Nachtraglich sind die
Herausgeber aber zu der Uberzeugung gekommen, dass das Zeichen
besser durch sva umschrieben sein wiirde, wobei v als Reprasentant
des labialen Halbvokals aufzufassen ware. Dann heisst es weiter, dass
derselbe Laut (., the same sound “) haufig in Verbindimg mit I, z.B.
Ipi (hes Ipa) 221 und selten in Verbindimg mit s, z.B. spa 238, erscheine.
* In der Tafel ist es durch em Versehen mit Ipi umschrieben ; in dem Zeichen
fehlt der i-Strich.
638 H. LDDEES —
Allein das ist eine petitio principii ; zunachst lasst sich doch nur
sagen, dass das subskribierte Zeichen in Verbindung mit s, s und I die
gleiche Form bat. Allerdings darf nicht iiberseben werden, dass das
Zeichen, wie es unter 221 in der von Eapson gezeichneten Schrifttafel
erscheint, keineswegs die typische Gestalt zeigt. Ich babe samtbcbe
in der Ausgabe und in Stein’s Ancient Khotan verdffentbchten
Eeproduktionen von Dokumenten daraufbin durcbgesehen und in
keinem einzigen Falle die Form der Schrifttafel gefunden. Zunacbst
ist die Schleife an der linken Seite der Vertikale im allgemeinen
schmaler als in der Zeichnung, bisweilen feblt sie ganz, so in Ipipeya 1,
Ipipe 17, Ipipeya 20, Ipipe 32 (samtUch Tafel I). Sie ist also offenbar
garnicbt beabsichtigt, sondern entstebt rein zufalHg, wenn die Feder
vom Ende der Vertikale des la wieder bocbgezogen wird, um den
Haken zu bilden, und nicht genau der Vertikale folgt. Das ist aber
nur eine kleine Verschiedenheit, auf die ich keinen Wert legen mocbte.
Wicbtiger ist, dass der Haken in alien Fallen viel weiter herunter-
gezogen wird als in Eapson’s Zeichnung und haufig noch einen
Schwung nach links zeigt, der bisweilen an die Vertikale heranreicht ;
vgl. ausser den angefuhrten Beispielen Ipipeya 164 (Taf. iii), Ipirnsu 164
(Taf. iii), palpi 165 Z. 7 (Taf. iii), Ipipahga 571 (Taf. ix) usw. Dadurch
gewinnt das Zeichen ein wesentUch anderes Aussehen als in der
Schrifttafel. Das gleiche gilt aber auch fiir die durch spa und s'pa
umschriebenen Zeichen. Auch hier gibt die Zeichnung der Schrifttafel
ein nicht ganz richtiges Bild, denn auch in diesen Ligaturen wird
der Haken stets nach unten und meistens dann noch nach links
gezogen ; vgl. spora 165 Z. 7 (Taf. hi), taniaspa 571 (Taf. ix), 580
(Taf. x), 581 (Taf. x), arispa 581 (Taf. x). Vom rein paliiographischen
Standpunkt aus wiirde sich also gegen die Auffassimg der Z-\ erbindung
als Ipa Oder Iva kaum etwas einwenden lassen. In einer so kursiven
Schrift, wie es die KharosthI ist, besteht aber stets eine starke Neigung
zm Anahnlichung urspriinglich verschiedener Formen, und meines
Erachtens konnen daher insbesondere bei Ligaturen Schliisse auf
den Lautwert aus der ausseren Gestalt des Zeichens nur dann als
giiltig angesehen werden, wenn sie sich auch sprachlich rechtfertigen
lassen. Das trifft aber in diesem Falle nicht zu ; die Lesung der
Ligatur als Ipa oder Iva scheint mir im Gegenteil zu sprachlich
unmoglichen Formen zu fiihren.
Nim begegnet uns die Ligatur allerdings fast ausschliesslich in
Fremdnamen, iiber deren sprachliche Form sich von vomeherein
nichts sagen lasst. Gliicklicherweise erscheint sie aber auch in ein
ZUR SCHEIFT UND SPRACHE DER KHAROSTHi-DOKUMENTE 639
paar Wortern, die aus dem Sanskrit stammen und daher einen
Riickschluss auf den Lautwert des Zeicliens enn6gliclien,und bisweilen
lasst sick ein solcher Schluss aucb aus Doppelschreibungen von
Fremdwortern zieben. In dem Yertreter von sk. kalydna wird das lya
meistens in der gewbhnlichen Form geschrieben, so in kalyanakari
mitra 499, 612, und in den Namen kalyanadhama 123. 560, 597, 601,
611, 618, gen. kahjanadhaftiasa 477, 619, 762. 763, und gen. guna-
kalyanasa 311. In 605 aber lesen die Herausgeber kalpanadhafna,
in 536 kalpanadhadiasa, in 756 gunakalpanasa, in 175 den Namen
kalpanad Sie sehen es als zweifellos an, dass hier der Wechsel eines
p-Lautes mit y vorliege, wie er auch sonst im Prakrit vorkommt.
Dass unter gewissen Bedingungen, die sich, nebenbei bemerkt, viel
genauer feststellen lassen als es bisher gescbehen ist, v fiir y eintreten
konnte, ist unbestreitbar ; dass aber kalydna jemals zu *kalvdna
geworden sein sollte, halte ich fiir ausgeschlossen ; es fehlt dafiir
an jeder Parallele. Wenn das fragliche Zeichen fiir lya eintritt, so
kann man daraus nur schliessen, dass es einen dem lya ahnlichen
Lautwert haben muss, und da es im Grunde nur in Fremdwortern
erscheint, so wird es nicht einfach eine andere Schreibung fur lya
sein, sondern ein der Fremdsprache eigenes palatalisiertes la. das man
gelegentlich auch in der Schreibung des Vertreters von sk. kalydna
verwendete. Da das subskribierte Zeichen in der Ligatur nach den
obigen Darlegungen offenbar eine kursive Form des ya ist, so mochte
ich vorschlagen, die Ligatur in libereinstimmung mit der sonst von
den Herausgebern gewahlten Umschrift modifizierter Zeichen durch
lya wiederzugeben, obwohl ein mouilliertes I den Laut wohl genauer
ausdriicken wiirde.
Anstatt des gewohnlichen lihita 649, lihida 144, 328, 437, 652,
likhida 331, 648, findet sich in 575 lyihida ,, geschrieben “. Da das
Auftreten eines Labials zwischen dem I und dem i natiirlich
unerklarlich ist, sind die Herausgeber gezwungen, das von ihnen
angenominene Ipihida als einen Schreibfehler anzusehen. Sobald
wir in dem Zeichen ein palatalisiertes I sehen, bietet sich eine viel
weniger gewaltsame Erklarung dar. Es zeigt sich, dass das lya seine
Hauptstelle in Verbindung mit i in den nicht-indischen Namen hat.
Im Index sind die mit lyi beginnenden Namen leicht zu iibersehen.
Es sind, wenn man von Varianten in der Schreibimg absieht, von
lyipaae bis lyivrasmasa nicht weniger als 23. Dazu kommen mit in-
oder auslautendem lyi : \kalyike'\ 757, kalyigeya 495, kalyigeyasa 207,
^ Ob in 351 Jcal.na zu kalyana oder kol^na herzustellen ist, ist ganz unsicher.
640
H. LUDEKS —
kalyigeyena ^ 207 ; kalyita 701 ; kalyisa 666 ; kilyigamciyana 164 ;
kolyige 93 ; kolyisa, kolyisa 8, 15, 29 usw., kol^sasa, kolyisasa 29, 110,
130 usw., kolyisasya 159, kolyisena 20, 53 ; tsulyita 74; mcdyigeya^
237, malyigeyasa 93, 277 ; molyina 131, 482, 581 ; sulyita 17 ;
aralyi, aralyiyasa 573 ; palyiya 596 ; pisalyiyami 122 ; alalyie oder
amalyie 406. Dieses massenhafte Auftreten des Zeichens vor dem
i-Vokal Uefert, wie mix scheint, die Bestatigung, dass es ein
palatalisiertes I ausdriickt, und wenn das lyi gerade ia den Fremdnamen
so haufig ist, so diirfen wir daraus den Schluss ziehen, dass diese
Neigung zur Palatalisierung in der Fremdsprache, was immer sie
gewesen sein mag, iliren Ursprung hatte.
In den auf das Sanskrit zuriickgehenden Wortern wird im allge-
meinen vor i das gewohnUche I geschrieben, so in den zahlreicken
Formen von likh (im Index von likhami bis likhidu und von lihafi
bis lihyati, in den abgeleiteten Formen von niskal, samkal und
panpalay, in denen auf die Stammsilbe ein i folgt (im Index
nikalitavo, nikalisyati und von nikkalita bis nikhalisyati, von
samyalitaga bis samgalidavya, von paripalitavo bis paripalidavya),
ferner in kamculi 149, 318, kamjuliyasa 343, pipali 702, silipatam 511,
priyasali ^ 83, 140, kamzavaliyana 725,* ebenso vor dem epenthetischen
i in muli, midiyami, tmdiyammi, muliyena (Index), Sicherlich stammt
aus dem Sanskrit auch vyalidavo mit den Nebenformen vyalidavo;
vyalidavya, vyalitavya, viyalidavo, viyalitavo, viyalidavya (Index),
ferner mit mehr oder minder Wahrscheinbcbkeit lisita 52, avalika
575, phalitaga 214, tmhuli ® 528, chagali pasu 613, khidini 349, und
der Name risaliae, visaKyae 722.* Aucb in Lehnwortern aus dem
Iranischen wird immer li geschrieben, so in dem schon zur Asoka-
Zeit ins Indische aufgenommenen livi in livivistarena, Kmstarena,
liviatararnmi (Index), in milima (Index), das auf p.eStp.eos- zuriick-
geht,’ aber ebenso wie satera, drachma uber das Iranische hiniiber
^ Text k.lpije vini.
- Text malpigeyo.
® Offenbar sk. priija-^yala ; warum ist hier i im Auslaut eingotrctcn ?
^ Der erste Be&tandteil des Wortea ist natiirlich iranisch.
* Mahiili kann aber nicht, wie im Index angegeben, = sk. makila sein, da dies
ein aus d entstandenes I hat, das in der Sprache der Dokumente als d erscheinen
musste ; siehc Festschrift W’ackemagel S. 306. Auch die Tdentifizierung mit pali
mahaflikd ist den Lauten und der Kedeutung nach nicht befriedisend.
® DaUmi 496 ist, wie im Index vermutet wird, wohl SchreibfehJer fur wilimu
ebenso diUk.yi 510 = sk. tUikfd fur didiksn. Unklar ist hali 83, livasa 109,
pfdiyrirnaga {‘i)
^ Thomas, Index.
ZUR SCHRIFT UND SPRACHE DER KHAROSTHI-DOKUMENTE 641
iibernoinnien ist, und wohl auch in Jcalihari 709, Mlihari 399d Es
lasst sich naturlich nicht entscheiden, wie weit die Schreibungen in
den echt indiscben Wortern die wirklicbe Aussprache wiedergeben, wie
weit sie historisch sind, Wenn aber, wie wir geseben, die Neigung zur
Palatalisieriing des I vor i in der Fremdspracbe bestand, so kann es
nicht auffallen, wenn gelegentlich einmal ein lyihida erscheint.
Ein zweites Beispiel fiir lyi anstatt li in einem Prakrit worte
liefert 162 : gJirita Idii 3 atrami prehidanui parupararivarsi ghrita
nasti hesa yo sesa hoati ahum-no pamcam-a varsa sarva galyiti sarva
gida. Galyiti ist unverstandlich. Ziehen wir in Betracht, dass der
Schreiber noch an zwei andern Stellen Silben ausgelassen hat — er
schreibt namake gya fiir naniakero arogya und tatnkami fiir tamka-
lamnii — so diirfen wir wohl annehmen, dass galyiti fiir samgalyiti,
das Gerundium ^ von samgal (sk. satnkal) verschrieben ist. Sanigal
wird haufig von dem Einsammeln von Korn und Wein, die als Steuer
abzuliefem sind, gebraucht. Die Ausdrucksweise ist in dem ganzen
Briefe nachlassig und unbeholfen ; was der Schreiber sagen wollte,
ist wohl : ,, 3 khi Schmelzbutter haben wir dorthin (zu euch) geschickt.
Von der vorjahrigen und der vorvorjahrigen Schmelzbutter ist
nichts riickstandig. Was riickstandig gewesen ist, das ist jetzt im
fiinften Jahre alles eingesammelt und alles empfangen.“
Ebenso begreiflich wie das Auftreten des lyi fiir li in indischen
Wortern ist umgekehrt aber auch die gelegentliche Schreibung li fiir
lyi in Fremdnamen ; so in lipe 754 gegeniiber dem ungemein haufigen
lyipe, lyipeya ; livarazmu 43 neben lyiparasma 102, lyivrasmasa 83 ;
lipu 80, 558 ; malina ^ 409 ; yalina 754 ; lalik. 701 ; tsugeli 642 ;
tsugaliya 170, tsogaliyasa 93 ; sugeli 650 ; yili 62, 259, 288. 701, yilika
642, yiliyasa 288 ^ ; simoliya 185 ; koliyammi 152 ; pisaliyade 64,
341, pisaliyammi 291, pisali . . . 351 neben pisalyiyami 122.®
Das lyi findet sich nun aber nicht nur in Eigennamen, sondern
^ Das Wort ist sicherlich im letzten Grunde eiiie Ableitung von sk. kalahak^ra
Streit machend das als kaUiharaa auch ins Sakische ubernommcn ist (Konow,
Saka Studies 140). Kahhari konnte wieder aus dem Sakischen zuruckentlehnt
sein ; das Auftreten des i bleibt in jcdem Fallo unerklart. Suhga 661 ,, aus
Suli stammend “ ist als Fremdwort anzusehen.
* Vgl. vajiti ,, gelesen habend “ in 152, 725, und 376, wo der Text vahti bietet,
sruniti 341.
® Nicht ganz sicher, vielleicht malena ; vgl. aber mohjina 131, 482, 581.
* Ein anderer Versuch, den Xamen zu schreiben, ist offonbar yilga 80.
^ Dass das i in dem mit lyi beginnenden Xamen sehr fluchtig gesprochen wurde,
zeigt die schon erwahnte Schreibung Ipanga 415 neben dem gewohnlichen lyipatiga
571 usw.
642 H. LUDERS —
auch in Appellativen. von denen wenigstens eins seiner Bedeutung
nach hinreichend klar ist, das ist palyi. Die Uberschrift des Tafeicbens
207 lautet : [aj}ii\yama avanammi s{e)sa palyi, „ der restliche joalyi
in dem Dorfe Ajiyama." ^ Es folgt eine allerdings niir sehr unvoU-
standig erhaltene Liste von Personennamen im Genitiv, hinter denen
die verschiedenartigsten Gegenstande genannt sind ; offenbar sind
es die Sacben, die jene Leute abzuliefern baben. Dann beisst es am
Scbluss zusammenfassend : p{i)mda sesa huda yam ca navaga yam ca
poranaga ghrida khi 10 4 4 asam[khai\ rajiya ^ 1 kosava 4 2 akisdha 1
thavamnae 4 1 kamxmta 10 4 2 goni 3 peda 3 pasu 1 rnasu xnilima 1 khi
4 1 pohgonena ^ amna militna 10 4 go 1, ,, in summa ist der Ruckstand
gewesen, sowobl der neue als aucb der alte : 18 khi Scbmelzbutter,
1 asamkharagiya (?), 6 langbaarige Decken, 1 oA-isd/ia-Decke, 5 Stiick
Tucb, 16 kamimta, 3 Sacke, 3 Kbrbe, 1 Scbaf, 1 milima 5 khi Wein
mit pongoha, 15 milim-a Korn. 1 Kub.“ Mir scbeint sicb daraus deutlicb
zu ergeben, dass palyi die Steuer ist, die das Dorf an das konigUcbe
Finanzamt abzubefern bat. Dazu stimmen die Angaben in 714,
einem Briefe des cozbo Takra an vasu Op^eya und Sfgaca : ahuno
esa tsugeta atm visajidemi ajiyama avanammi palyi dhahiasa praceya
yahi adehi puniga ajiyama, ai'anammi palyi cimtidaga ghrida pasava
kosava arnavaji thavastae raji narnmatae cdindri kammamtana amha
maka ogana crofna amha siidae kamamta * yarn ca aniha palyi sarva
spura agela lyipeya tsugetasa ca hastamnii cavala isa visajidavo „ jetzt
babe icb Tsugeta dortbin (zu eucb) gescbickt wegen der gesetzlicben
Steuer in dem Dorfe Ajiyama. Wie von dort friiber die auf dem
Dorf Ajiyama (begende) Steuer festgesetzt ist, Scbmelzbutter, Scbafe,
langbaarige Decken, arnavaji-Decken, Teppiche, raji, Filzdecken,
^ Im Text steht [ajhil^ama und s.sa, Ajhiyama ist wahrscheinlich falsche
liCsung, vielleicht K^chreibfehler fur ajiyama, das sechsmal in den Texten erscheint.
2 So nach dcra Index zu lesen.
® Da die Zeichen fur 1 und na sich sehr ahnlich sehen, ist sicherlich statt des ini
Texte stehenden poiigohe 1 jxmgone.na zu lesen ; vgl. pofigohena milima 1 masu in
574 (dreimaJ), masu pongonena pariatammi milima 1 khi 4, masu khi 4 pongohana,
khi 4 masu gida poiigohena, masu prahuda presidavya khi 3 poiigoiiena, masu pongohena
khi 3 prahvda anitama, masu nikhasta milima 1 khi 4 pofigohena, masu pongohena
khi 10 2 anitamti in 637. Pongoha ist offenbar eine Woiterbildung von ponga, das m
225, wiederum in Verbindung mit masu erscheint; isa pir.ci ginidemi masu pohga 2.
Der in der Ausgabe S. 314, Anm. 2 vermutete Zusanimenhang von pohga mit dem
Xamen poniga bcsteht auf keinen Fall, auch mit poiika (oder protsa) in 317 hangt
das Wort kaum zusamraen. Ponga scheint ein bestiramter Behalter fur Wein zu sein,
pongoha ist vielleicht alles was zur Aufhewahmng des Weins in einem solchen gehort;
das Suffix erinnert auffallig an das sakische Suffix -iiha, -auha.
* Ausgabe : su^ ekamarrifa.
ZUE SCHKIFT UND SPRACHE DEE KHAEOSTHI-DOKUMENTE 643
camdri foi«(»mim<a>!a,^ferner maKa, ogana, cronui. ferner suRae kamamta,
und was es sonst an Steuern gibt, alles das ist vollstiindig in der
Obhut des a^eta Lyipeya und des Tsugeta scbleunigst bierher zu
schicken.“ Aiif diese Steuerabgabe von Ajiyama geht ferner 275 ;
yahi purvika adehi ajiyama avanammi samvatsari palyi cirntitaga
visati matra varsa hutamti eda palyi atremi achimnidetha yahi eda
kilamadra atra esati pratha eda palyi imade pravamnaga praJiidatna
tena pravamnagena eda palyi cavala lepaia yatma agetasa ca cavala
sarva spara — isa lisajidavo, „ wie friiber vor dort (bei euch) die
jahrlicbe auf dem Dorf Ajiyama (liegende) Steuer vor zwanzig
Jahren ^ festgesetzt ist, diese Steuer habt ibr gerade dort ausgesetzt.
Wenn dieser Keilbrief dort ankommen wird, . . . diese Steuer.^ Wir
haben von bier eine Anweisung gescbickt. Auf diese Anweisung
bin ist diese Steuer scbleunigst (in der Obbut) * von Lepata und des
yatma ageta scbleunigst ganz und vollstandig bierber zii senden.“
Eine Reibe von Dokumenten betrifft den palyi in deiu Dorfe
Peta. 165 ist ein Brief des ogu Kirtisama an den cozbo Kranaya und
den sothatngha Lyipeya. Nacb den einleitenden Floskeln schreibt er ;
avi peta avanammi palyi paruvarsi sesa yam ca imavar.ii pmlyi taha
sarva spora tommihi sadha isa visajidavo yati tade puritna pacima
visajisyatu pamthammi parasa bhavisyati tuo sothamga lyipeya tanu
gothade vyosisasi, „ weiter : der Rest der vorjahrigen auf dem Dorfe
Peta (liegenden) Steuer und was die diesjabrige Steuer ist, sind ganz
und vollstandig sofort zusaminen bierber zu scbicken. Wenn da von
(etwas) vorber (und etwas) nacbber gescbickt werden sollte (und) es
unterwegs geraubt werden wird, so wirst du, der sothanxga Lyipeya,
(es) aus der eigenen Farm bezahlen.“ Es folgen Bemerkungen liber
den Ankauf von Schmelzbutter und die Mahnung, den palyi der vega
kihni Frauen, der in Korn besteht,® vollstandig zu scbicken. Dann
^ Camdri kamamta auch 272, vielleicht Silber-Arbeiten. Das fia hinter karjimamta-
weiss ich nicht zu erklaren.
^ Die Worte visatimatra varsa hutamti sind doch eher zu cirntitaga zu ziehen ab
zu achirnnidethaj da nicht anzunehmen ist, dass das Dorf zwanzig Jahre lang keine
Steuer entrichtet haben sollte.
^ Hinter palyi scheint etwas ausgelassen zu sein.
* Hinter agetasa ca ist wohl hastammi weggefallen. Auch die Wiederholung von
cavala beweist, dass der Brief nachl&ssig geschrieben ist.
^ Von dem le^a kilmi striyana palyi wird auch in 211 gesprochen : vega kilme
striyana palyi na anisyarriti, und weiter in 714: avi vega kilme striyana palyi spara
pruchidavo. Von dem vega kilme dhama, von dem auch die Entrichtung des palyi
abhangt, ist in 481 die Rede. Der Ausdruck vega kilme erfordert eine besondere
Untersuchung.
644
H. LUDERS
fahrt der Schreiber fort : avi palyi uta teneva sadha isa visajitavo
ma imci tomgana jmride uta vithisyatu tasa uta praceya raya saksi
lihidaga Icridaga livistammmi anatilekha atra gada tahi cozbo kranayasa
lihami eda karyami tuo cita kartavya esa lyipeya na cita kareti, „ auch
das Steuer-Kamel ist mit diesem zusammen bierher zu schicken.
Auf keinen Fall soil seitens der totngas (die Sendung des) Kamel(s)
verzogert werden. In betreff dieses Kamels ist ein Konigszeugen-
Dokument angefertigt worden. In ausfubrlicher Darstellung ist ein
Befeblsschreiben dorthin (zu euch) abgegangen. Ich schreibe dir,
dem cozbo Kranaya. Um diese Angelegenbeit musst du dich k iimm ern.
Dieser Lyipeya kummert sich nicht (darum).“ Das Steuer-Kamel ist
offenbar eine Zusatzsteuer fiir das Dorf Peta. Die Angabe, dass
in betreff dieses Kamels ein raya saksi lihidaga gemacht ist, lasst
darauf schbessen, dass iiber die Yerpflicbtung, dieses Kamel zu liefern,
zwischen der konigbchen Regierung und den ortbcben Behorden
Meinungsverscbiedenbeiten bestanden. Das ,, Konigszeugen-Doku-
ment “ wird doch wabrscheinlich ein Dokument sein, in dem die
Aussagen von Zeugen zu gunsten der koniglichen Anspriiche
protokolliert waren.
Wenn uns auch der in 165 erwahnte anatilekha nicht erhalten ist,
so haben wir doch in 42 einen kilamudra, in dem praktisch dieselbe
Forderung gestellt wird : yahi purvika adehi peta avanammi sarnvatsari
palyi cimditaga paruvarsi palyi sugnutana anada pieavida yahi eda kila-
mudra atra esati pratha edapalyi[ya] praceya vasu lyipeya [pra]. .davo
sarva spara cavala agita sarngapeyasa hastami isa lisajidavo yahi
purvika adehi peta avanammi sarnvatsari palyi uta cimditaga se uta
na vrdhaga na krisaga siyati teneva palyiyena sadha uta isa visajidavo
sesa palyi syati sfara visajidavo grida palyi purva cavala prahadavo,
„ wie friiher von dort (bei euch) die jahrhche auf dem Dorfe Peta
liegende Steuer festgesetzt ist, (in der Hohe) ist die vorjahrige Steuer
dem Sugnuta . . . ^ zu ubergeben. Wenn dieser Keilbrief dort (bei
euch) ankommen wird, ... ist wegen dieser Steuer der vasu Lyipeya
zu befragen (?).^ Sie ist ganz und vollstandig schleunigst in der
Obhut des agita Samgapeya hierher zu schicken. Wie friiher ist von
dort (bei euch) als eine jiihrliche auf dem Dorfe Peta (liegende) Steuer
ein Kamel festgesetzt. Sollte dies Kamel nicht alt (und) nicht mager
^ Der Plural sugnutana ist auffallig. Anada verstehe ich in diesem Zusammen*
hange nicht.
* Die in der Ausgabe vorgeschlagene Erganzung zu prochidavo ist nicht ganz
s'lcher.
ZUE SCHEIFT UND SPEACHE DEB KHAEOSTHI-DOKUMENTE 645
sein, so ist das Kamel zusammen mit jener Steuer hierher zu schicken.
Sollte ein Steuerrest sein, so ist er vollstandig zu schicken. Die
Schmelzbuttersteuer ist schleunigst zuerst zu senden.“
Wahrscheinlich bestand die Verpflichtung jahrlich ein Kamel als
Steuer zu Uefern auch fiir andere Dorfer. In dem konigUchen Briefe 70,
der nach der Aufschrift trasa avanammi eine Angelegenheit des Dorfes
Trasa betrifft, lesen wir : yahi purvika adehi rnalbhayasa vamti ^
cauravarsi palyi vithidaga migacasa ^ vamti caura uta vithidae yahi eda
kilamudra atra esati pratha eda palyi keti vithidaga syati dhaciyasa
hastami isa visajidavo ,, die wie friiher von dort (bei euch festgesetzte)
Steuer ist vier Jahre lang bei Malbhaya zuriickgehalten worden.
Bei Vugaca sind vier Kamele zuriickgehalten worden. Wenn dieser
Keilbrief dort (bei euch) ankommen wird, ... ist diese Steuer, soviel
(davon) zuriickgehalten sein sollte, in der Obhut Dhaciyas hierher
zu schicken." Es liegt jedenfalls die Vermutung nahe, dass die Zuriick-
haltung der vier Kamele mit der vier Jahre langen Zuriickhaltung
der Steuer in Zusammenhang steht.
Ich habe palyidhaina oben in 714 durch „ gesetzliche Steuer “
wiedergegeben ; der Ausdruck scheint dem deyyadhamym nachge-
bildet zu sein, das im Pali im Sinne von reUgioser Gabe verwendet
wird. Dass palyidhama jedenfalls die konkrete Steuer bezeichnet,
geht deuthch aus 164 hervor, einem Privatbriefe des Lyimsu an
seinen Vater, den cozfco Lyipeya, wo von der Einsammlung und
Absendung des palyidhama die Eede ist : avi ca ahono isa peta
avanemciye palyidhama prace sutha vihedemti paruvarsi pago pake
palyi kida avi kilyigamciyana palyi prace vihedernti yati eta lekha
atra esati cavala palyidhama, isa visarjidavya yo puna arnha adehi
rajade samarena tsamghina kve^namdhina palyidhama isa mama
prochamti avasa cavala samghalidavya isa prahadamja avi ca avasa
pago isa visarjidavya, „ und ferner : Jetzt drangen sie hier ® die
Einwohner des Dorfes Peta sehr wegen der gesetzlichen Steuer. Fiir das
vorige Jahr hat Pago die paArc-Steuer entrichtet. Auch wegen
der Steuer der Kilyigarnciyas drangen sie. Wenn dieser Brief dort
(bei euch) eintreffen wird, ist schleunigst die gesetzliche Steuer
hierher zu schicken. Da sie hier ferner die andere von dort, von
der Provinz (zu entrichtende) gesetzliche samarena-, tsamghina-,
kvemarndhinaStevLej: von mir fordern, so ist sie auf jeden Fall
* Text hier und naehher amti.
* Text vagacasa.
’ Der Brief scheint aus der Hauptstadt geschrieben zu sein.
646
H. LXIDEES —
schleunigst einzusammeln (und) hierlier zu senden. Auch ist auf
jeden Fall Pugo hierher zu schicken.“ Auch in 211, einem Privat-
brief, kann palyidhafria nur etwas wie „ gesetzhche Steuer “ bedeuten :
tanu gothade palyidhaivM achinasi ainnesa palyi na pragada nikhaksi —
iuo atra trifi nvahatra si avagajena imade hemamtammi palyi praceya
lekJia risaji{de)mi triti bJiagade eka bhaga na palyi isa visajidesi yati
ahuno bJiuya eda palyi na spora isa anisyaniti nacirena tuo ima varsa
vasammi isa agamisyasi, „ die gesetzliche Steuer von dem eigenen
Gute enthaltst du vor, die Steuer der andern gibst du nicht offen
heraus. . . . Du bist dort der dritte hohe Beamte. Bei Gelegenheit
habe ich von hier ini Winter wegen der Steuer einen Brief geschickt ;
du hast die Steuer auch nicht zu einem Teil vom dritten Teil * hierher
geschickt. Wenn sie jetzt fernerhin diese Steuer nicht schleunigst
hierher bringen, wirst du in kurzem in diesem varsavasa hierher
komnien.” Dass zwischen palyidhama und dem einfachen palyt
kaum zu scheiden ist, zeigt auch der Vergleich von 714 palyidhafnasa
pricha hoti, „ es findet Einforderung der gesetzlichen Steuer statt
mit 725 avi kamzaraliyana palyiyasa anada pricha ganana kartavo
spura isa visajiddvo „ auch ist die . . . Einforderung und Ziihlung
der Steuer von Dingen, die an die Schatzbeamten zu liefern sind,
zu machen (und sie) ist vollstandig hierher zu schicken
Von palyi ist in den Dokurnenten noch ofter die Rede, ich kann
hier abet nicht auf die verschiedenen Arten des palyi eingehen, da dafiir
zunachst die Bedeutung von Ausdriicken wie kilmeciya, samarena (?),
tsantgina, koyimamdkina usw. klargestellt werden miisste. Auch das
Verhiiltnis. von palyi zu harga und suka soli hier nicht behandelt
werden. Schon aus dem Angefiihrten scheint mir zur Geniige
hervorzugehen. da.ss palyi die von einzelnen Personen wie von der
Provinz, dem Dorf und anderen staatlichen Verbiinden an die konigliche
Regierung in Naturalien zu entrichtende Steuer ist. Hochstens in
450 konnte an eine Abgabe an eine Privatperson gedacht werden.
Da .schreibt Lyipana an ein Ehepaar, Kroae und Lugaya, nach den
einleitenden Floskeln ; an ca vasanitammi atra krisicatra kararnnae
ma itnci avakasa karisyata ahuno caturtha varsa huda mahi palyi
achinasi yo tahi atra yotha bhmnaksitra taha vikrinamnae parihara
^ Triti bkarjade eka bhaga scheint ein idiomatischer Ausdnick fiir ,, wenig “ zu
sein. Ein ahniicher Aiis<Iniek findet sieh in dem gleichen Zusammcnhang in 3l'> ;
yatha pUTiika adehi tsamgina palyi cinididaga taha ardhnde ardha na anemti ida bahu
d^ramnaga huamti, ,, Wie fmher ist von dort die f^awffi'na-Steuer festgesetzt. Sie
gen nicht die Halfte von der Haifte. Sie sind hier viel schuldig.“
ZTJE SCHEIFT UND SPRACHE DER KHAROSTHI-DOKUMENTE 647
odidemi tahi saniadue bharyae putra dhidarehi isa agamdavo isa hrisivatra
kartavo mahi palyi sudha rotanma avi curaina sa ^ isa anidavo amTia
palyi mahi na kicamaga, „ und ferner ; im Friihjalir soil keineswegs
Erlaubnis gegeben werden, dort (bei eucb) das Land zu bestellen.
Jetzt ist es das vierte Jahr gewesen, dass du mir die Steuer {oder
meine Steuer) vorenthaltst. (Was) die Kuhfarm (betrifft und) das
Ackerland, die du dort hast, so babe ich (dir) freigestellt (sie) zu
verkaufen. Du musst niit deiner Mutter, deiner Frau und deinen
Sohnen und Tochtern hierher kommen. Hier ist das Land zu bestellen.
Mir ist als Steuer nur Krapp ^ und aucli curania hierher zu schicken.
andere Steuer brauche ich nicht.“ Es ist aber sehr wohl mbglich,
dass auch hier palyi in dem gewohnhchen Sinne gemeint ist, indem
Lyipana die zu liefernden Dinge als palyi bezeichnet, well er sie selbst
an das konigliche Steueranit abzuliefern hat ; jedenfalls haben Krapp
und curama, ofter curo>ita, crofna, einen Platz in den koniglichen
Steuerlisten ; siehe 357, 387, 714 usw.
Mit der richtigen Lesung und der Feststellung der Bedeutung
scheint mir auch die Ableitung von palyi gegeben zu sein : es wird
nichts weiter sein als sk. ball, „ Steuer, Abgabe Da in den
Dokumenten ofter eine anlautende Media durch die Tenuis wiederge-
geben wird,^ macht die Identifizierung von palyi und bali im Grunde
keine Schwierigkeit. Jene Schreibungen treten aber doch immer
nur sporadisch auf, wahrend in palyi das p ebenso ausnahmslos
erscheint wie das ly. Das liisst doch darauf sclihessen, dass man sich
des indischen Ursprungs des IVortes nicht mehr bewusst war, und ich
mochte es nicht fiir unmoghch halten. dass bali in der bereits zu
palyi veriinderten Gestalt aus einer Sprache wie etwa dem Tocharischen.
wo b zu p werden niusste und wahrscheinhch auch das I vor i
palatalisiert wurde. entlelint worden ist.
Das zweite Wort, in dem ein ly vor i erscheint, ist vyalyi, das
offenbar mit ryala oder riyala zusammenhiingt, da beide Beiworter
von uta ,, Kamel sind.'* In 437 wird beurkundet, dass Kornpala
und sein Sohn Sugiya ein kudi Miidchen fur 45 verkauft haben. Die
^ Text ciirama.‘in.
2 So nach Burrow, BSOS. 7, 787.
^ Z.B. kaiii dramgatnmi, tamdn, tita, tivase, tivirOy tosa, trakhma, tramghanu, tritkoy
poga usw.
* In 703 wird riyala ohne deutlicbe Beziehung auf ein Kamel gebraucht : isa
srawamiin anamdasenasya riyala krita atra garhanaf's es hindert aber niehts auch
hier zu libersetzen : ,, hier hat er ein r/i/a/a-Kamel des sraman-a Anaindasena gekauft,
um dorthin zugehen'b
648
H. LUDEKS —
Kaufer haben aber nicht die ganze Summe gezahlt : taha eta kampala
sugiya sa ca ducaparisa muliyami viyala uta 1 padichitamti tade sesa
muli 3 vithitaga Jiuati, „ dieser Kompala und Su^ya haben ein
viyala-K.am.el im Werte von 42 erhalten ; der von dieser (Kaufsumme
verbleibende) Rest im Werte von 3 ist vorlaufig nicht bezahlt worden."
590 ist ein Kaufvertrag Tiber eine Frau Lyipaae. Ein gewisser Sarnca
hat sie an den Schreiber Ramsohka verkauft : tivira ratnsohlcasa
paride stri lyijaaae muli smnca gida eka uta viyala capxirisa muliyena
padichita bhiti uta akra trim muliyena tavastaga 1 hasta 10 2 bhiti
tavastaga hasta 10 1 amha sutra muli gida 4 4 sarva jnmda muli hoti
20 20 20 20 10 4 4, ,, von dem Schreiber Ramsohka hat Sarnca als
Preis der Frau Lppaae erhalten : ein viyala-Kamel im Werte von
40, ein zweites Kamel, (namlich ein) akra (!), im Werte von 30, einen
12 Ellen langen Teppich, einen zweiten 11 Ellen langen Teppich. Ferner
hat er als sidra-Preis 8 erhalten. Der ganze Kaufpreis betragt in
summa 98.“ Ich bin iiberzeugt, dass das unverstandliche akra nur
ein Schreiberversehen fiir akratsa oder amkratsa ist, das in den
Dokumenten bfter als Beiwort von uta auftritt. In 569 wird als
Zahlung fiir ein Ziehkind ein akratsa-Kame\ angegeben (kuthaksirasa
uta akratsa ditaga). Nach 195 haben eine Anzahl von Leuten, die
gemeinsam ein Opfer veranstaltet haben, von vasu Opgeya em
aOTl-raisa-Kamel fiir dieses Opfer geholt (vasu opgeyasa paride uta
1 amkratsa yaynhami nitamti). 330 ist eine Urkunde in betreif von
a^raisa-Kamelen (akratsa utana prace), in 383, einer AufsteUung
der lebenden und der toten Kamele der koniglichen Stuterei, wird
auch ein amkratsa erwahnt, ebenso in der durch Bruch verstiimmelten
Tafel 428 (uta akra . . .) und in 163. wo der Zusammenhang nicht
klar ist. Den Beweis fiir die Richtigkeit der Verbesserung von akra
zu akratsa scheint mir vor allem 592 zu hefern. Es ist ein Vertrag
zwischen Pulnamto und dem Schreiber Ramsohka uber den Kauf
eines Am^i-Madchens. Der Preis, den Pulnarnto empfangt, ist auf
ein a/.Ta^sa-Kamel und eine Khotan-Decke festgesetzt, wobei das
aA:ra<sa-Kamel genau so wie in 590 mit 30 bewertet wird (muli
uta 1 akratsa trim muliyena pulnamto padichida arnha anga muli
khotani kojava 1). Vergleicht man die Preisangaben in 437, 590 und
592, so ergiebt sich, dass ein viyala-Kame\ um ein Drittel teurer war
als ein arnkratsa-Kamel und zum teil noch hoher bewertet wurde.
Anderseits ist ein viyala-Kamel etwas weniger wert als eine
vierjahrige Kamelstute, wie 420 zeigt : korara kamjaka ari-saraspasa
vyala uta 1 dharanaga huati yam kala ari-saraspa ichita maramnaya
ZUR SCHRIFT XIND SPRACHE BER KHAROSTHI-DOKUMENTE 649
tarn liolammi ari~saraspa kamjake svasu sariyae hastammi krita
uta pruchamnae ahuno kamjaka uthita sariya srivamrnasa ca catuvarsi
uti 1 vyosita tena kamrarnna utvaramrsi ditaga prace puhgetsa ^ 7
arohaga muli 4 2 kamjaka patama nita, ,, korara Kamjaka war
Ehrwurden Saraspa ein I’l/akt-Kamel schuldig. Als Ehrwiirden
Saraspa sterben wollte, da hat Ehrwiirden Saraspa es in die Hand
(seiner) Schwester Sariyae gelegt, das Kamel bei Kamjaka einzu-
fordern. Jetzt ist Kamjaka vor Gericht erschienen. Er hat eine
vierjahrige Kamelstute an Sariyae und Srivarhma ausgeliefert.
Aus diesem Grunde, weil er ein im besseren ^ Alter stehendes (Tier)
gegeben hat, hat Kamjaka ein puhgetsa als den iibersteigenden
Preis von 6 wieder herausbekonmien.‘‘
Ein puhgetsa muss wiederum eine Art von Kamel sein. In 561
heisst es, dass nach dem Urteilsspruch des Richters in einer Diebstahls-
angelegenheit : jimoyasa dazasya paride amkratsa puhgetsa odarasya
ca nidavya huati. Ich kann dem Text einen Sinn allerdings nur abge-
winnen, wenn statt odarasya ca vielmehr odara sa ca zu lesen ist :
„ von dem Sklaven Jimoya waren ein amkratsa, ein puhgetsa (und) ein
odara wegzuholen.“ Das ist nicht geschehen : puhgetsa odara atremi
vithidae, ,, der puhgetsa (und) der odara sind dort (bei euch) zuriick-
gehalten worden Um einen ahnlichen Fall handelt es sich in 359.
Da klagt eine Person, deren Name verloren ist, dass ihm im koniglichen
Gerichte durch Gerichtsbeschluss zwei puhgetsa-Yia,me\e als Busse
von seiten des Pgina zugesprochen seien. Von denen sei ein Kamel
gegeben worden, das zweite Kamel sei nicht gegeben worden : pginasa
paride puhgetsa uta 2 vyochimnidae tade eka uki dita biti uta na denati.
In 401 handelt es sich um einen Streit um die Miete fiir ein Kamel ;
als Miete fiir dieses Kamel war ein puhgetsa zu liefern : eda utasa
parikre puhgetsa nidavo. 526 berichtet von der Aussage eines
gewissen Sugnuta : sugnuta vimhaveti yatha edasa kuhaniyarnmi
kalu kuvaya amna gida puhgetsa utena eda uta na visajesi, ,, Sugnuta
zeigt an, dass kalu Kuvaya in Kuhaniya von ihm Korn fiir ein
pu/(^rfsa-Kamel gekauft hat. Dieses Kamel schickst du nicht.®
Mit derselben Sache beschaftigt sich noch ausfiihrlicher der Brief
530 : ahuno isa sugnuta vimhaveti yatha edasa kuvayena amna gida
^ Text hier und stcts j>uhgehha mit puiigetm in der Note. Ich halte pungetsa im
Hinblick auf amkratsa fur die wahrscheinlichere Lesung.
- Ich nehme an, dass utvara fur uitara steht.
® Es ist nicht klar, wer der Angeredete ist. Hinter na ist ein Stuck von der Tafel
abgebrochen und daher vielleicht etwas von dem Texte weggefallen.
VOL. vni. PARTS 2 AND 3.
42
650
H. tUDEES —
isa hihaniyammi milhna 3 puhgetsa uta midi abhisaniitamti fade
uvadae baJiu varsa hiitatnti na denati, „ jetzt zeigt Sugnuta hier
an, class Kuvaya hier in Kuhaniya von ihm 3 milima Korn gekauft
hat. Sie haben sich liber ein pu>igetsa-^ame\ als Preis geeinigt.
Seitdem sind viele Jahre vergangen, es ist nicht gegeben worden
Es muss auffallen, dass ein so geringer Wert wie 3 milima Korn hier
einem Kamel gleichgesetzt sind, aber es steht daniit im Einklang,
dass in 420 ein puiigetsa-Kamel auch nur mit 6 bewertet ist. ^\ir
konnen daraus den Schluss ziehen, dass puhgetsa der Ausdruck fiir
ein ganz junges Kamel ist, und dafiir spricht auch eine weitere
Bemerkung in 530. Nachdem dort befohlen ist, die Sache zu unter-
suchen, heisst es : bhudartha eva habsati uta varsaga nacimti dadavo ^
athava amna ayogena dadavo, ,,sollte es sich in Wahrheit so verhalten.
so ist ein jiihriges Kamel . . . zu geben ® oder es ist Korn mit Zmsen
zu geben. “ Da das jahrige Kamel bestimmt ist. auch den Schaden
zu ersetzen, der durch die jahrelange Nichtbezahlung des pungetsa
entstanden ist, so muss es inehr vert sein als ein puhgetsa.^
Wir konnen also gewissermassen eine Preisliste fiir Kainele
aufstellen : 1) catuvarfi uti 46-48 ; 2) viyala 40-42 ; 3) amkratsa 30 .
4) varsaga ; 5) puhgetsa 3 milima-6.* Da ferner gesagt ist, dass ein
).'dyrt/n-Kamel an Jahren hinter einer vierjahrigen Stute und ebenso
ein puhgetsa hinter einem jlihrigen Kamel zuriickstehe. so wird es
vahrscheinlich. dass die drei unbekannten Ausdrticke viyala. amkratsa
und puhgetsa sich auf verschiedene Altersklassen beziehen. Ich nibchte
auch annehmen, dass viyala eben.so wie sicherlich amkratsa und
puhgetsa aus der Fremdsprache stammen, iiber deren ISatur wir
vorliiufig nichts wissen. Jedenfalls kann ich der S. 318 geiiusserten
Ansicht der Herausgcber, dass vyala, viyala mit sk. rydila, vydla.
zusammenhange, nicht beipflichten. Dagegen spricht die Bedeutung,
denn das Sanskrit Wort dient im allgemeinen nur zur Bezeichnung
von Raubtieren und Schiangcn, und wenn es im Sinne von ,. tiickisch.
boshaft *■ auch von einem Elefanteii gebraucht wird, so kann, wie
der Zu.sammenhang zeigt, in den Dokumenten doch unmoglich von
^ Text varsagann cinitidadaro^ mit der Bemerkung, dass fiir dm auch rci, dhini
oder rdhi gelesen werden kann.
^ yadmti oder, wie die 8illx*n sonst zu lesen sind, verstehe ich nicht. Moglich
ist es naturlich auch, dass varsagana zusammengehort und ein cimti {‘t) von den
jahrigen gemeint ist.
® Die in 383 dreimal wiederkehrenden Worte taya {taya) dhitu piingt^tsa verstehe
ich nicht.
* Es komnien aber auch abweiohende Preisangaben vor. 8o wird im 571 von einem
zweija.hrigen Kamel, das 50 wert war, gesprochen {uta 1 duvarsaga pamcasa muliyerm).
ZUR SCHRIFT TJND SPRACHE DEE KHAEOSTHI-DOKUMENTE 651
tiickischen “ Kamelen die Eede sein. Ausserdem konnte I'ljada.
I'yala in dem Prakrit der Dokumente nur als *vyalJa, *viyai}a erscheinen,
da tier altes d niemals als I auftritt.
Zu vyala gehort min als Femininum injalyi. das sick in 591 und 546
findet. Der Text von 594 ist ganz fragmentarisch. Nacli dem Satze
avi vasamtammi iitasa karnmna lain vamti jalpila, .. ferner : im
Fridijahr hat er mit dir wegen des Kameles gesprochen ”, sind nur
noch die Worte erhalten : amsa ah(ii)i>o i[cha\ vuta liynhji. aus denen
sich nicht viel entnehmen liisst, zumal anstatt icha auch iiia oder
idha gelesen werden kann. Xur soviel ist klar, dass vyahji sich auf
ein Kamel bezieht. Leider ergibt sich auch aus 546 nicht viel mehr.
Der sramana Dhamapriya schreibt : wahi idi dvi vynlyi 1 I'ugeynm
yothadarana paride niyidavo huati. ,, ich hatte von den Leuten auf
der Kuhfarm des Vugeya zwei Kamele. eine vyalyi zu holen.*' Im
folgenden ist aber immer nur von einer Kamelstute {uti). die
Dhainapriya von dem Sohne des Vugeya erhalten hat. die Rede,
und es lasst sich nicht entscheiden. ob damit eine der beiden nicht
naher bezeichneten Stuten oder die vyalyi gemeint ist. IVahrschoinlich
findet sich vyalyi ein drittes Mai in 341 : avi ca atra rnyaka klinla
Ida yo trevarsa yarn ca tade a [tkavar-f .] . . [ga] ede ida -fotliamya kohjisa
cahna.iasa ca hastarrd isa visajidavo ckadasi mamsya dasnrnrni pisaJi-
[yarnmi) sarnga{lidavo) ^ — hotu yn [a» da vyalya vyala'] idae na ^
irnci isa visajidavo pisaliyade pmlhama [. .] isa rirnnatilekha pralia-
davo ede sniniti pisaliyade isa risaji{da)i'od Statt vyalya vyala wird
vyalyi vyala zu lesen sein : .. und ferner : Die Kamele des kbniglichen
Gestiites dort (bei euch), die dreijahrigen und die in hoheren Jahren
sind (?) ^ als die.se. diese Kamele sind in der Obhut des sotharnga
Kohusa und des Calmasa hierher zu schicken. Sie sind am zehnten
des elften Monats in Pisaliya zu sammeln.^ . . . die vyalyi und vyala
Kamele sind keineswegs hierher zu schicken. Yon Pisaliya i.st zuerst
ein Brief mit der Anzeige hierher zu senden. Diese (Kamele) sind
(erst), nachdem man (die Antwort) gehort hat. von Pisaliya hierher
zu schicken.” Bei der Unsicherheit der Textiiberlieferung lasst sich
vorliiufig auch aus dieser Stelle kaum etwas Genaueres fiber vyalyi
erschliessen.
Das Auftreten des hj ist nicht auf die Stellimg vor i beschrankt ;
^ Die Erganziingen stanimcn von niir.
- Text uta e nu.
Aiharar.^ . . . ga ist unsioher. Das tade lasst darauf schliessen. dass ein Wort mit
komparativer Dedeutiing wie etwa *ajhivar8aga oder *utaravarsngn folgte.
^ Die aiif eine Lucke folgenden Worte hotu yo nu da verstehe ich nicht.
652
H. LUDERS —
ly findet sich auch vor a, e und o, und zwar wiederum am haufigsten
in Fremdnamen : alyaya 9, 370, 575, 709, alyayasa, alyayena 214 ;
alyasena 684 ; Icilyagi 348, kilyagiya 322, kilyagiyasa 216, 322, 348,
kilyagisa 36 ; kilyama 246 ; kolyaya ^ 701 ; kolyarasa 701 ; pulya
37 ; lyaka 701 ; lyakdha 701 ; spalyayasa ^ 579 ; spalyaya ® 709,
spalyaya 506 ; solyaka * 701 ; calyeya 596 ; kalyotsa ® 585. Ein
paarmal erscheint ly vor a imd o auch in Appellativen, aber abgesehen
von dem oben bebandelten kalyana nur in Wortern, die nicht oder
wenigstens nicht direckt aus dem Indischen stammen.
In 151 findet sich fiinfmal gilyamyasa, gilyamyam oder gilyamya,
gilyamya pasava (oder pasu). Da der Ausdruck mit ghritasa oder
ghrita pasava (oder pasu), ghriti pasava (oder pasu), palyi pasava (oder
pasu) in derselben Tafel auf einer Stufe steht, scheint er eine besondere
Art von Schafen zu bezeichnen.
In 318 wird in einer Liste von gestohlenen Dingen zweimal
ein lyohnana genannt : samimna citraga lyohnana und sujina kirla
lyohnana. Von den Beiwbrtern ist vorlaufig nur citraga klar, da aber
lyohnana zwischen kamculi „ Jacke “ und prahuni „ Gewand “
eingeordnet ist, wird es vermutlich ein Kleidungsstiick bezeichnen.
Vollig unklar ist sulyagamdha in dem Satze avasa suly/againdha
prahadavo, mit dem der Brief 127 schliesst. Der eigentliche Inhalt
des Briefes ist verloren, und es bietet sich daher kein Anhaltspunkt
fiir die Bestimmung der Bedeutung des Wortes dar.
Besser steht es in dieser Beziehung um silyoka, silyoga, das in
den Dokumenten bfter erscheint. In dem zum teil schon oben
behandelten kbniglichen Schreiben 359 heisst es ; aid garahati yatha
edasa isa rayadvaranuni silyoka lihitaga pginasa paride puiigetsa uta 2
vyochinmidae tade eka uta dita biti uta na denati atra (sa)muka anada
pruchidavo yatha silyoganimi Khidaga imthuami ahcmo yatha dhamena
vibhasivo ® na {im)ci tatra amhatha kartavo „ ferner klagt er, dass ihm
hier im kbniglichen Gerichte ein silyoka geschrieben ist : es sind
(ihm) zwei puiigetsa-lxa.me\e (als Busse) von seiten des Pgina
zugesprochen worden. Von diesen ist ein Kamel gegeben worden,
das andere wird nicht gegel)en. Dort (bei euch) ist in Anwesenheit
die Sache zu untersuchen. Wie in dem silyoga geschrieben ist, genau
so ist jetzt nach dem Gesetze zu entscheiden, nichts ist in dieser
^ Text kolaypa (I)ruckfehler).
® Im Index avalpaya.
* Oder kalyatsa.
2 Vielleicht spalyiyasa,
* Oder whika, wie im Text.
® Sehreibfehler fiir vibhasidavo.
ZUK SCHEIFT UND SPEACHE DEE KHAEOSTHI-DOKUMENTE 653
Sache abzuandern.“ Hier ist silyoka, silyoga deutlich ein Schriftstuck,
das das Urteil des Gerichtshofes enthalt. Dieselbe Bedeutung
bat silyoka in 312 : ahuno isa jihmaya garahati yatha edasa catata
ayasa ca paride mamnusa Kamki parikraya vyockimnidaga silyoka
likidaga kitae, „ jetzt klagt bier Jihmaya, dass ihm der Lohn fiir den
Mann Karnki von seiten des Catata und des Aya zugesprochen worden
ist. Es ist ein geschriebenes silyoka gemacht worden Es folgen die
gleichen Weisungen fiir die weitere Behandlung der Angelegenheit
wie in 359. Auch in 561 bezieht sich die Bemerkung silyoka lihidaga
kidae auf das Urteil, das der kitsayitsa Luthu und der cozbo Kamci
in dem Prozess des Sunamda wegen gewisser gestohlener Sachen
gesprochen haben, und auch hier wieder wird fiir die weitere
Behandlung der Sache bestimmt, die Entscheidung nach dem
friiheren Spruche, wie das silyoka geschrieben ist, zu treffen : yatha
kitsayitsa luthu cozbo kamci sa ca hastama vyochimnidaga silyoka
lihidaga syati tena vidhanena ya(tha) dham(e)na nice kartavo. In 729,
wo nur der Schluss des koniglichen Schreibens erhalten ist, ist in
die Phrase noch hasta lekha, offenbar „ handschriftlich “ eingefiigt :
yatha silyoga hasta lekha lihidaga siyati tena vidhanena nice kartavo.
Dass das silyoga aber nicht unbedingt den Urteilsspruch enthalten
muss, geht aus 492 hervor : savathena saksiyena rayakadhamena
jyruchidavo atra na paribujisatu saksiyena savathena matralekhami
lihidavo sa ca silyogena hastagada isa visajidavo, ,. (die Sache) ist mit
Eid imd Zeugenvernehmung nach dem koniglichen Recht zu unter-
suchen. Sollte sie dort (bei euch) nicht klar gestellt werden, ist es
mit der Zeugenvernehmung und dem Eid in einen matra-Bnef^ zu
schreiben und dieser mit dem silyoga in Obhut hierher zu
schicken “. Da es sich hier uni den Fall handelt, dass kein Urteil
gefallt wird, so kann silyoga hier wohl nur das Protokoll der
Verhandlung sein. Und das Gleiche gilt auch fiir 471, wo die
Auseinandersetzung eines Streitfalls, von der nur die Halfte erhalten
und die daher schwer verstandlich ist, mit den Worten schUesst :
eda prace vistarena silyoga lihidaga rayadvarainmi prahadavo ede
vevatuga hastagada rayadvarainmi visajidavo, ,, wegen dieser Sache ist
ein ausfiihrlich geschriebenes silyoga an den koniglichen Gerichtshof
zu senden. Diese Prozessierenden sind in Obhut an den konighchen
Gerichtshof zu senden “.
1 Matra ist wohl so viel wie mantra ; ein matralekha ware also ein benachrich-
tigender Brief. Die I,esung ist aber nicht sicher ; es kann auch rnaha gelesen werden.
65-1
H. LUDERS —
Silyoga hat aber eine noch allgemeinere Bedeutung. la 140 handelt
es sich um Korn, das Kupsimta, dem Schreiber des Briefes, gehort
und sich zmn teil in andern Handen befindet ; yo tade amnade
nikhastaga amnesa ditaya sarva silyoganimi kritaga, ,, was von dieseni
Korn abgeliefert, andern gegeben ist, das ist alles in ein silyoga
gemacht Der Ausdruck silyogatnmi kritaga kann bier kaum etwas
anderes bedeuten als ,, in einer Liste urkundlich aufgezeicbnet
Im Sinne einer urkundlichen Liste wird silyoga auch in 470 gebraucbt,
wo aber im Einzelnen Unklarheiten bleiben : ahuno isa yapgu
vimnaveti yatha paruvarsami esa ida nikJialida asya inm varsami
ganana katvetha silyoga likitaga katvetha eda silyoga isa a[nida }iuda\
„ jetzt zeigt Yapgu bier an, dass er im vorigen Jahr ein Kamel beraus-
gegeben hat. Dieses habt ihr in diesem Jabre mitgezahlt.^ Ihr baht
ein geschriebenes silyoga gemacht. Dies silyoga ist bierber gebracht
worden
Neben silyoga, silyoka findet sich einmal, in 582, aucb sulga. Die
Tafel entbalt einen Kaufvertrag uber einen Acker. Wegen dieses
Ackers ist spater ein Streit entstanden, und die Entscbeidung in
dem deswegen gefiihrten Prozesse ist als Nachschrift dem Vertrage
angefiigt. In dieser Nachschrift stebt der Satz : eta sulga lihidaga
pramatia huda, „ dieses geschriebene sulga ist giiltig gewesen “.
Mit dem sulga ist offenbar die Kaufurkunde gemeint. und sulga kann
daher nur als eine nacblassige Schreibung ftir silyoga angesehen werden.
Die Bedeutung von silyoga ist somit ., urkundliches Scbriftstuck “,
und ich mochte glauben, dass das Wort nichts weiter als sk. doka
ist. Das mag in anbetracht der Bedeutungsverschiedenbeit sebr
kiibn erscheinen. wenn auch innerhalb des Sanskrit die Bedeutungen
von sloka ,, ScbaU, Gerausch, Euf, Kuhm, Strophe ” recbt weit
au.seinander gehen. Formell wurde der Einschub des Teilvokals obne
weiteres begreiflich sein, da wir im Pali siloka, im Prakrit siloga,
siloa haben. Auch im Sakischen erscheint sloka als silo. Aber auch
das auffallende linguale .s und die Palatalisierung des I hat ihre
Parallele in der Arsi-Sprache. Toch. Gramm. S. 60 wird angegeben,
das.s sk. sloka sieben mal als slok, acht mal als slyok erscheint, allerdings
immer in der Bedeutung „ Strophe Dass zwischen diesem .slyok
und unserm .nlyoka ein Zusammenhang besteht, scbeint mir
unverkennbar.
^ Die Cborsetzung ist unsicher. Statt asi/a kann tana gelesen werden. Kah'ftha
(las nur hier vorkommt, seheint 2. Plur. Praet. von f.rzu sein.
^ l)ie Lesung anida huJa i-st nicht sicher. Der iSchluss (ies Briefes fehit.
ZUR SCHEIFT TJND SPEACHE DEE KHAEOSTHI-DOKUMENTE 655
Der Palatalisieriing des I vor i steht eine Palatalisierung des n
vor i in den Fremdnamen zur Seite, somamciini, acuniya, acuniyasa^ ;
apnigJiade, aptiiya, apniyani, apniyasa ^ ; kunita, kunitasa, kohita,
koiiitasa ; kunisae ■, krinilasa ; ninieyasa ; senima, senimnia ;
tnanigeya. Dass das i sehr fluchtig gesprochen wurde, beweisen
Nebenformen wie mahgeya, mangeyani. In Wortern indiscben
Urspmngs und in iranischen Lehnwortern, auch in dem Ortsnamen
mna^ bleibt das n vor i unverandert,® und auch in Fremdnamen
wird ofter ni geschrieben ; siehe anisisge ; kenika, kenikasa, keniga,
kenigani ; cinika, cinikani, einikasa, ciniga, ciniya ; nanimasanisa ;
nivagasa ; pamniyasa ; panimcaade, panicanade ; pumniyade ;
ponigana, poniganasa, ponigani ; ponicga ; bumni, bmnnvmci, bunini-
yammi ; manigi. Die Yerteilung von ni und ni ist also eine ganz
ahnliche wie die von lyi und li.
Zu der Palatalisierung von I und n vor i stimmt schliesslich auch
der Vorschlag eines y vor anlautendem i, der, wie aus dem Index zu
ersehen, in den Fremdnamen ausnahmslos zu Tage tritt und gelegent-
lich auch auf indische IVorter iibertragen wird, so in yima ’237 neben
dem gewohnlichen ima, yiyopravaninaga 348, 416 neben iyo pravamnaga
655 usw. Der Ansatz eines palatalisierten I (lya) scheint mir danach
nach alien Seiten gesichert zu sein.
[Das Manuskript dieser Arbeit war vor Erscheinen des Aufsatzes
von T. Burrow. " Tokharian Elements in the Kharosthi Documents
from Chinese Turkestan,” JRAS., 1935, S. 667 ff. abgeschlossen.]
^ In 327 in der Schreibung aciinoyam.
^ In 251 einnial auch apnemsa.
^ Der Xame dhamhila in 678 ist ofiFenbar von dhdnya abgeleitet ; ein Bruderdes
Dhamnila heisst ebenda Dhaninapi'da. Unklar ist vorlautig krisaga nohi in 383.
Iranian Elements in Khowar
By G. Morgenstierne
T^HOWAR, the predominant language of Chitral and of the adjacent
parts of the Gilgit district, is characterized on the one hand
by a tenacious preservation of ancient lA. sounds, forms, and words,
and on the other hand by the existence of a remarkably large number
of foreign elements. According to Sir George Grierson, Khowar “ in
some essential particulars agrees rather with the Ghalcha languages
to the north And, drawing attention to the fact that the Chitral
valley ^ was formerly inhabited by Kalashas, he expresses the opinion
that the originally homogeneous Dardic population of Kafiristan,
Chitral, and Gilgit “ was subsequently split into two by a wedge of
Kho invasion, representing members of a different, but related,
tribe coming from the north [of the Hindukush] In whatever way
one may be inclined to interpret the position there can be no doubt
that Kho., when compared with the neighbouring Dardic dialects,
presents many pecuUarities which deserve our attention.
Among the words included in the list given by Sir George ® in order
to exemplify the difference between Kho. and other Dard and Kafir
languages some may be of lA. origin. But it is none the less remarkable
that Kho. should differ so widely from its neighbours, even as regards
a number of the most common words, e.g. such as denote parts of the
body. Thus, corresponding in most cases to words of lA. origin in
Shina and Kalasha, Kho. has yec ‘ eye ’, /t'pnq ‘ mouth ’, li'gini
‘tongue’, ri'gis ‘beard’, ca'mot ‘finger’, do'yur ‘nail’, A-raw, nrga
‘back’, us'gdr ‘lung’, qhol ‘bone’, is'ki (q1) ‘heel’, is'qi ‘jaw’,
bliq ‘ pupil ’, 'zdnu ‘ knee ’, bazu ‘ arm ’. Some of these words come
from Ir., others are of unknown origin.*
The Ir. loanwords in Kho. which form the subject of this article
are very numerous. They may roughly be divided into the following
groups, which, as will be seen, cannot always be clearly distinguished :
I. Loanwords from (Modern) Prs.
II. Loanwords from some Middle Ir. language.
III. Loanwords from the Pamir dialects.
lY. Loanwords from some undefinable or unknown Ir. source.
^ LSI., viii, 2, p. 133.
^ Or, at any rate, Lower Chitral.
® Loc. cit.
* Cf. also, for instance, rin ‘ wine 'reni ‘ dog ’, is'kini ‘ silk etc.
658
G. MORGENSTIERNE
I
The first group is by far the largest one, but also the least interesting
from the linguistic point of view. Most of the Prs., or Ar.-Prs., loan-
words in Kho. belong to that well-known host of “ Islamic ” words
which has swept victoriously over half of Asia. They need not detain
us here. But there are a few which present phonetic or semantic
peculiarities of some sort and which call for some brief remarks.
It is noteworthy that in some words Prs. voiceless plosives have
been rendered by Kho. aspirates. Thus ; phost ‘ skin Icham ‘ palate
ba'than ‘ native country A secondary aspiration appears also in
some Kho. words of lA. origin, for instance in phan ‘ palm of the
hand ’. But in Badakhshi Prs. there is a tendency towards the
aspiration of voiceless plosives, and it is possible that this is the reason
why Kho. has chosen the aspirate in some cases. We may compare
the situation in Ossetic, which possesses two series of voiceless plosives,
and may employ either of them to present foreign ones.
Ba'than is remarkable also on account of its b corresponding to
Prs. w. This word cannot have been borrowed before the Kho. change
of r- > b-, and I can find no reason w'hy the w- w'as not retained, just as
in ivaxt, ivazlr, etc.
A curious form is sabli'kl ' a kind of horse fodder, lucerne ’. It is
evidently borrowed from Prs. cf. saftal, sautal ‘ lucerne ’ ; but the
exact form from which it is derived is unknown.^ But there must be
some connection between the Kho. form and ‘ Sanskrit ’ sai-pi-li-
h'ie ( < sak-b'ji-lhk-ka) given by Li Shi Chen.®
gurzen ' garden ’, which from Kho. has penetrated into Kal. and
Pal., is possibly a cross between *gurzel, *gulzer < Prs. gulzdr and
Prs. guUan.
The ■■ inidla which has been assumed for the hypothetical Prs.
dialect-form *gulzer, appears also in other Kho. words. Thus we find
ke'lPc (Kal. ku'l'iic) ‘ fathom ’ < Prs. guide, pe'lesk ‘ rug ’ < Prs.
palds, 713 ' u'es ‘grandchild’ < Prs. nawdsa, kiteb, kitep (Werchikwar
kitep, Yidgha ketiu), but also kitdb ‘ book ’ < Prs. kitdb, sexik
^ In yhi'ran ‘ shirt ’ < Prs. pirahan the aspiration has been transposed. The
origin of phincdl, pilu'dl ‘trousers ’ is unknown to me. It rhymes with Prs. Barwdl
< mlicdr.
^ Palola (Dardic dial, in Chitral) sabrlki probably comes from Kho. Another kind
of lucerne is called sau'thnl in Pal., sau'tM m Gawar-Bati, and saz'th-tl in Kalasha.
^ t ide Laufer, JSijio-Irauira, p. 214.
* Cf. Horn, Gr. Ir. Ph.^ i, 2, 33.
IRANIAN ELEMENTS IN KHOWAR
659
(O’Brien) ‘ powder horn ' < Prs. sax{ak). Cf. also Munji rikeb ‘ stirrup ’
< Prs. rikeb = rikdb.
Kho. nivesik ‘ to write ’ agrees with the common E. Prs. form
which has introduced the s of the preterite into the present stem.
But I am unable to trace the exact source of such forms
as riwis ‘ rhubarb ' : Prs. riwds, Psht. raivas ; naxoi ‘ pea ’ :
Prs. nuxLid ; yalwanj ‘ kite ’ : Prs. yaleu'dj, etc. ; angdh ‘ awake ’ :
Prs. dgdh ‘ aware ’ ; yuzdr ‘ footstep, trace ’ : Prs. guhdr ' passage ’.
krenj (Kal. krin) ‘ wrinkle ’ corresponds to Prs. gurinj, but a k is
attested in Bal. kirisk, Kurd, kurisk < *kurinck.
A further form peculiar to Kho. (and the adjoining Werchikwar)
is tu'vek ‘ gun, musket ’ : Prs. tufang, tiifak.^
Finally it may be mentioned that Kho. has borrowed from Prs. a
few words which I have not come across in any of the neighbouring
languages. Thus, e.g. : guddz ‘trouble, grief’ and iskdU ‘excuse’,
probably from Prs. sikdl ‘ cunning, deceit ’, cf. sikdlis ‘ care, roguery
IskdU also exemplifies the tendency to render Prs. si-, su- + a
consonant by Kho. is-. Thus : iske'na ‘ awl ’, iskama ‘ belly ’, ispel
‘ whistling ’,^ istdr ‘ horse ’, istiyln ‘ sprouts ’ (Psht. teyujia), if from
Prs. sitey ‘ straight ’. The same tendency prevails in the development
of lA. words in Kho. and it is impo.ssible to decide whether the initial
ist-, etc., in the loanwords is of Kho. or of dialectical Prs. origin.
As regards istor it is possible that it has not been borrowed from
Modern Prs., where the word is hardly ever used, but from middle
Ir., cf. e.g. N.W. Turfan Phi. ‘siivr.
This leads us on to the next group of loanwords.
II
In an earher publication ® I have tried to show that Kho. contains
a number of other “ Middle-Ir.” loanwords. It is, however, impossible
to draw a clear line between words of Middle E.Ir. origin and such as
may have been borrowed from one of the Pamir dialects at an earlier
stage of development. But, at any rate, some of the Kho. words of
^ The distribution of the words for ‘ gun ’ : miltiqt banduq, and the various
forms assumed by tufang in Ir. and lA. would require a special study. Here it suffices
to note that tuvek must have come from the north-west. The Dard dialects of the
Kunar valley have forms borrowed through Psht. tjtdpak. Most Pamir dialects (and
Ormuri) have the Turki word miUiq, or they employ the old local word for ‘ bow ’
with a new meaning.
2 But Kal. sispfe ' whistling ' ; i4pfVriy Kati 6ph ‘ flute ’ are genuine.
® Report on a linguistic mission to Afghanistan, p. 75 f.
660
G. MOEGENSTIERNE —
“ Middle-Ir.” appearance probably come from a language of civilization
and not from a neighbouring local dialect.
Among the most important of the words dealt with are harm
‘ mirror ’ < *dhm ; caxur ‘ spinning-wheel ’ : Sogd. caxr ‘ wheel
etc. ; nimez (Kati namoc) ‘ prayer ’ : Phi. Tiarndc, Kurd, nimez, etc. ;
zar ‘ poison ’ : N.W. Phi. zakr, Kurd, zdr ; duwarth ^ ‘ door ’ :
Old Pers. duvardi- ‘ portico ’ ; Yrdtiu ‘ pregnant ’ ^ : Sogd. yirdn
‘ heavy ’. Also xatan ‘ house, room ” must be an early loanword from
Ir., cf. Gabri xada ‘ house and Slav, xata ‘ hut ’ which is probably
also borrowed from Ir.®
On the other hand, it is very doubtful if chat, cat (Kal. chAt)
‘ lake ’ can have been borrowed from Ir. (Av. cdt- ‘ well ’), as suggested
loc. cit.
But there are some other words which are certainly of early (W.
or E.) Ir. origin. Thus we find ro'ci ‘ a fast ’ : Prs. roza, Psht. rdza,
Brahui loanword roca ; anic ‘ forehead ’ ; cf. Bal. ariisay
< *dmcaka * ; ca'tir ‘ tent ’ (borr. from Kho. into Kal. and Yd.) :
Prs. cddar, -ir ; girwa'loy ‘ peach ’ from an earlier form of Prs.
gurddlu with substitution of rw for VS®; -zeri (hi jrusi-zeri ‘ kitten ’)
< -*cidriya-, cf. Av. azi-cidra-, V9hr1c6-c° ‘ descended from a snake,
a wolf ’ ; Ujpdrzing ‘ battle-axe ’ : Prs. tahar-zln(g), \Vkh. tapdr, etc. ;
*hozik ‘ to play (chess, etc.) ’ * : Prs. bdzani, bdxtan ; -yon ‘ like,
resembling ’ ; Sogd. ydn ‘ maimer, kind ’, cf. Pashai goni ‘ like ’,
Hiirkin (N.E. Caucasian) -yuna ‘ like ’. Also the interesting word
'droxvm ‘ silver ’ must have passed through Middle Ir. on its way from
Greek to Kho.
It is also tempting to derive Kho. pax'turi ‘ the sunny side of a
valley or a hill ’ from Av,, Phi. apdxtar{a-) ‘ northern ’ with u from
unstressed a (cf. asur ‘ he is ’ < *dsati) and a secondary suffix. In
Upper Chitral the earhest known home of the Klios, the sunny side of
the valley would be the north-western, in Mastuj the northern one.
^ Kal. durwM, durgat may be adaptations of the Kho. word.
* It is noteworthy that the native lA, word gar'b'tn is used about animals, while
the foreign and nobler one is reserved for human beings.
® This derivation is rejected by Meillet {£SIj,t 28, p. 133), who does not, of course,
know the Kho. and Gabri forms. It is not necessary to assume, with Jokl. firw*S^.,
12, p. 66) that xata has passed through an Ugrian language with x- < A-. Cf. also
Yidgha-Munji xeyo ‘ wall ’ ?
* Vide yrs,, 5, p. 40.
® Ir. S becomes r in Kho. But in the group *rS this treatment was excluded.
® Pres. 1 pi. bnjisif pret. 3 pi. bozitanij etc. A different word is bozik ‘ to divide ’
of lA. origin.
lEANIAN ELEMENTS IN KHOWAR
661
A much more doubtful case is toxmiran which O’Brieu and also
mj Kalasha informant give as the Kho word for daughter-in-law.
It has a decidedly Ir. appearance, and one could easily imagine a
Middle Ir. *tdxme-8dn denoting the daughter-in-law as the continuer
of the family hneage, the " receptaculum seminis ".
But Colonel Lorimer, who has kindly undertaken to ascertain the
exact meaning of the word for me in Gilgit, informs me that it does
Tiot mean daughter-in-law, but “ is a term used of two sets of parents-
in-law vis-d-vis each other, and is used by them as a term of address
to each other This seems to upset the otherwise plausible derivation
of toxmiran.
It is perhaps conceivable that the meaning given by Lorimer is,
after all, a secondary one. If the word did originally denote
the daughter-in-law, it may have got its sphere of meaning extended
so as to include her parents, and subsequently have lost its primary
signification and have been restricted to its present use.
We do not know from which Middle Ir. dialect or dialects Kho.
may have borrowed. There exists, however, possibly one indication
regarding the source of some of the words mentioned above. The
Kalasha name for the Khos is Pdtu, which can be derived from
*Partaw and may originally have been applied to a group of
Parthians who found their way to Chitral and established their rule
among the native tribes of the country.
Ill
From early times down to the present day the inhabitants of
Chitral have been in constant and close contact with the Ir. tribes
to the north and west, and their dialects have been influenced by Kho.
This is especially the case with Yidgha, which has been spoken in
Chitral for many generations. But Wakhi also, although being
separated from Chitral ^ by a high range of mountains, has submitted
to the influence of Kho. to the extent of borrowing even some personal
pronouns from this language.-
But Kho., on its side, has not been averse to adopting a number of
words from these numerically and culturally unimportant Ir. border
tribes. We do not know hmv the process of borrowing has taken place,
' The existence of recent Wakhi settlers in the Yarkhun valley does not alter the
situation in this respect.
2 Cf. Rep. Afgh., p. 79.
662
G. MOKGESSTIERNE
but we must remember that the circumstances favour an extensive
linguistic exchange. Small groups of settlers are constantlv emigrating
from one valley to another, peasants cross the passes to trade in the
nearest bazaar or in search of temporary employment as labourers,
shepherds meet in a friendly or unfriendly way on the grazing grounds
in the mountains. The village communities are small, and the local
dialects have no great prestige. The result is a lack of backbone,
a linguistic lability which offers little resistance towards the intro-
duction of foreign elements.
But, even so, it is remarkable that Kho. should have borrowed
from Waklii, or from an earlier form of that language, some of the
most common words. ^ Thus we find
Kho. do'yur ' finger-nail ’ < Wkh. 'ddger, di'^r, etc. < *n(a)g- (cf.
Capus’ Wkh. ‘ hindiger ’) < *nak- with Wkh. k < x.
Kho. ri'gis ‘ beard ’ < Wkh. regik, rayls < *fra-gaisa- (acc. to Skbld).
Kho. ispril ‘ fruit-blossom ’ < *spruy : Wkh. spray, spray. Cf. also
Kho. is'prenzik, isprenpk ‘to blossom ’< Ir. *sprenja{ya)-,
cf Sogd. 'spry- ■ to blos.som Sak. Jiaspaljs- (caus.).
Kho. isprask. isprdzg ‘ ceiling ’ < *spdrsk ( ? ), Wkh. si'pask, Shgh.
s'pdsc ■ rafter '.
Kho. sonthu ' raven ’ < Wkh. sond, cf. Oss. sunt.
Kho. pdz ■ breast ’ < Wkh. piiz. Sar. puz, etc.
Kho. roxhl ' embers ' < Wkh. raxni(g) ‘ fire ’.
Kho. yarik, yerik ' to turn (into) '. etc. < AVkh. yir- ‘ to turn round '.
Kho. pot ‘ polo-ball ’ < AA’kh. tup ‘ playing ball '.
Kho. sa'puk ‘ hoof ’ (AA^erch. sapa). AA'kh. has the Prs. loanword sum,
but there may have existed an earlier genuine AA'kh. form with
the characteristic AA’kh. change of/ > p (cf. Av. snfa-).
Kho. u'ciru ‘rain’, zomeik ‘to yawn', sareik ‘to appear, seem’
correspond to AA’kh. u'ur zim ‘ yawn ’ ; suhi'iy- ‘ to appear ', but
these words may have been l)orrowed bv Kho. from some other
Ir. dialect (cf. O.ss. zambin ‘ to yawn ’).
And in the ca.se of Kho. xel (Kal, khel) ' perspiration '. an'ddu
' fever '. way ' price, money ’ : Wkh. Jfil, anduv. way, which can
scarcelv be genuine AA’kh. forms, it is ecpially possible that both
languages have borrowed from a common Ir. source. Nor is it certain
that AA’kh. kirzepc. kizipci ' magpie ’ i.s the form from which Kho.
kisipi has been borrowed. Cf. Shgh. kixejw, AA’erchikwar yasep. etc.
^ Cf. Report on a linguistic mission to Afghanistan, p. 76.
IRANIAN ELEMENTS IN KHOWAE
663
The following words, which are common to Kho. and Wkh.,
have an Ir. appearance, but they have not been traced in anv other
Ir. language and their origin is therefore uncertain : Kho. n'xdr :
Wkh. ica'xdr ‘ ladder ’ ; Kho. biz'bdr ' eagle, kite ’ : Wkh. bispiir
(*biz ( 1 ) + par ■) ; Kho. san'gur (Kal. sdhgr'ur'i) ‘ guts, entrails ’ :
Wkh. hngar ; Kho. jjx, zox ‘ thorn, bramble ’ : Wkh. zax, zax.
On the other hand, the vocalism and also the isolated position of
the Wkh. word render it probable that AVkh. tiirt ' ford ' is borrowed
from Kho. thiirt. Kho. naslc ' beak ‘ corresponds to Wkh. niick, Sar.
niisk, but can scarcely be separated from Gawar-Bati nasiik, Dameli
nas, Pashai nosuk, ndhtk.
It is surprising that Kho. should have borrowed any words from
Yidgha, the dialect of a small and unimportant border-community.
Yet this seems to be the case, although the situation is considerably
obscured by the circumstance that Yd. has evidently borrowed many
words of Ir. origin from Kho., and it is often difficult to distinguish
such words from those which may have been adopted into Kho.
from Yd.
The following are the most probable examples of Kho. borrowing
from Yd., or an older form of that dialect : —
Kho. hii'ddm ‘ limb, body ’ : Yd. hadntue < Av. handdman-, with
regular change of ml < d. h- appears in Yd. without any regard
to etymology.
Kho. yeik, yeim ' to spin ' : Yd. yliini. with regular loss of inter-
vocalic Cf. Munji ylwum, Wkh. Hip-. Sgl. giw-, yei'- < *gciip-.
Kho. ivrazni (also Pal.) " pillow ' : Yd. virzane < *varznai
<., *barznika-. But Ktio. w-irzidlm ' bolster ' < *mrz{d)iri is an
earlier loanword and is probably in its turn the source of Yd.
vrazi'dine.
Kho. xu'rk ' wild rose ’ : Yd. axrlo < *d-xari(d- ( ? ). cf. Prs. xdr. etc.
Kho. pezanik ‘ to entrust ' : Yd. piznm-, Munji p^joH- < *pati-
jamaya-.
Kho. hlnj ■ needle ’ : Yd. sinjo, sunp < *sdziio, cf. Mj. Hzno. The
Kho. word can scarcely be connected with Kal. sks. suzlk. etc.
Kho. fros ' muzzle, lip of an animal ' ; Yd. fruse < Av. Jrasa-
' forwards ' A -ka-. It is not certain that the Kho. word is
borrowed from Yd.
Kho. dreik ‘ to pour out. .sow, throw, shave ’ : Yd. drl- with the same
varieties of meaning. The word is also found in Munji and must
be Ir. (*Jian-drdtvaya- ?).
664
G. MORGENSTIERNE
Kho. vrenjeik ‘ to fry ’ : Yd. vroc- (for -waj- < *vrin]- ?). Cf. Bal.
brijag.
Kho. ju-saxa ‘ a period of two years Only given by a Yidgha as
the translation of his own loh-saxa. If correct °saxa < *saxw-
may be connected with Ir. sale- to pass the time, vide Alphab.
List, s.v. SOOK-.
Kho. trispoy ‘ dandelion ’ : cf. Yd. trisp ‘ sour ’.
Kho. ro'min (O’Brien ro'menhu) ‘ aspen ’ : Yd. ra'meno. Also in
Mj. and therefore scarcely of Kho. origin.
Kho. usturik ‘ to flee, run, jump ’ : Yd. ustusc- ‘ to jump ’. Apparently
the original Yd. present stem was *ustur-, *ustur-, or *ustur-,
ustusc- being the ancient past stem.
Kho. au'zetu ‘ heifer (two years old) ’ : Probably < early Yd. *avze9o
< *abi- (or *upa-)-zq0yd- ; cf. vazaxo ‘ pregnant animal ’ <
*upa-{ ? ) zqdya-, p9Zexi ‘ male sheep one to two years old ’
< *paHzqdyaka-.
Kho. carar ‘ spark ’ corresponds to Yd. carox, but similar forms are
found also in other Ir. languages. Yd. yerndnu ‘ whirlpool,
eddy ' is borrowed from Kho. yerndnu, yerddnu (Kal. ger'dhan
fr. Kho.), in its turn a loanword from Ir., cf. Prs. girddb, Sgl.
ydrhdv and especially Waziri Psht. yrhnbai < *gartdna- ?
The nature of the relationship between Kho. ims ‘ bridle ’, ziryeik
‘ to lament, cry ’, sayurj ‘ large hawk %vdy ‘ price ’, and Yd. yutviz,
zirly-. sayurs, huy ^ is unknown. Kho. trus'ni ‘ thirsty ’ is probably
a loanword from Ir., but A^d. truhie ® is borrowed from, and not the
source of, the Kho. word. I do not know whether Kho. tresna'yuli
‘ thyme ’ is of Ir. origin. Also Yd. tanou ‘ rope ’ (Biddulph) is probably
borrowed from Kho. tanau < Ir., cf. Prs. tanudan ‘ to twist ’.
Finally we may mention that Kho. nol ‘ rushes ’ is derived from
Skr. nada- and not borrowed from Munji nal.
The third Ir. Pamir language bordering upon Chitral is Sanglechi-
Ishkashmi. The strongest evidence of its influence upon Kho. is
Kho. u'drz ‘above’: Sgl. vardz {<*varz- <*barz-). Kho. u'sanu
‘ whetstone dumik ‘ to card wool ’, yxrdik ‘ to turn round ’ remind us
of Sgl. vasin (< *av9sdn), d»mb-,yort- ‘to walk about’ (early loan-
word from Phi.). But it is by no means certain that the Kho. words
^ Palola loanword sayiirz.
^ Cf. above, p. 662.
’ Also Yd. tisrw, etc., of Prs. origin.
IRANIAN ELEMENTS IN KHOWAR
665
come from Sgl. Cf. e.g. Yd. afseno, Prs. afsdn ‘ whetstone ’ ; Yd.
lib- < *dumh- ‘ to card etc.
If Kho. a'drax ‘ hillside, forest ’ is connected with Ishk. alax,
dlax ‘ hill ’ ( < *ardaxa-) the borrowing must have taken place at an
early date. Kal adraJch, Pal. adrax ‘ hillside ’ may easily be loanwords
from Kho., from which these languages have borrowed very freely.
Kho. cid‘m ‘ brass kettle ’ (> Kal. ci'dhin) recalls Ishk. ciidan ;
but both words are probably of Prs. origin, cf. caudan, cu° ‘ cast metal '.
h'dan ‘ kettle ’ is found also in Yazghulami. Finally may be mentioned
the very remote possibility of the curious Kho. word yo'U ‘ hare ’
having been borrowed from a lost (or not recorded) Sgl. *yoli <
^gausiya- ‘ (long-) eared ’. But what about Kal. g^'ava'rialc, gdva'nak
‘ hare ' 1
IV
IVhile the Dardic languages south of Kho. contain a number of
loanwords from Psht., such words are extremely rare in Kho. It is
not till quite recently that the two languages have come into contact,
Kho. expanding towards the south and Psht. towards the north in the
Kunar valley and Dir.
But there are many words in Kho. which have a general E.Ir.
appearance, or which agree more or less with forms found in several
Pamir dialects. Such words will be found in the following Alphabetical
List (cf. s.vv. cojnk, last, menu, muryulum, pdlitm, rost, soxcik, sd,
sapik, war, werkhu, wrazun,xdl, {a)xomik, xora,zdy, zcrc) which includes
also, for the sake of reference, all the words mentioned above, and a
number of doubtful loanwords, or such as could not easily be fitted
into the groups dealt with above. There are many other Kho. words
which contain un-Indian sounds, but which have not been found
either in Iranian, Turkish, or Burushaski. But there is always the
possibility of new words having been created within Kho. itself at a
time when foreign sounds had already been adopted into the phonetic
system of the language. Cf. e.g. yot ‘ dumb ’ and dag ' boy which at
the same time contain cerebrals and the un-Indian sounds y and q.
In Psht. and Bal. words of a similar mixed type are quite numerous.
At any rate, I hope that the list, in spite of its shortcomings and
its inconclusiveness, may give some impression of the extent to which
Kho. has been subject to Ir. influence, and that other scholars may be
able to trace the source and history of many words which I have had
vot. vni. PARTS 2 ANn 3 . 43
666
G. JIORGENSTIERNE —
to leave unexplained. I need not add that niv Kho. material is bv
no means exhaustive.
Alphabetical List of Ira^iiax Words in Khowar
dra'roy ‘ eructation ’. With a kind of reduplication from Prs. drdy ?
Yd. ara'roy is borrowed from Kho.
anic ‘ forehead v. p. 660.
atiddii ‘ fever v. p. 662.
awgdih) ‘ awake v. p. 6.59.
anzeik ‘ to send < Ir. *han-zaya-, cf. Av. zd{y)- ‘ mittere ’ ?
mzoxt (Lor.) ’ waste, desolate ’. Of Ir. appearance, but without
any known connection. Sgl. zenz- : zuy6 ‘ to lift, seize ’, etc.,
cannot be compared.
is'klm ‘ silk v. p. 657.
is'kena ‘ awl v. p. 659.
x'sanit ‘ whetstone ’, v. p. 664.
iskurdi ‘ short '. Cf. Kurd, kurd (v. Bartholomae ZDMG., 50, 702) ?
isprii ‘ blossom ’, isprenjik, Up° ‘ to blossom ’, v. p. 662.
isprdsk ‘ ceiling v. p. 662.
■tti'gdr ■ lungs ', cf. Prs. saxdr.
istari ‘ star ’. Prob. lA., cf. Ashkun is'ta, etc.
istor ‘ horse v. p. 659.
iskdli ‘ excuse ’, v. p. 659.
iskama ‘ belly ' v. p. 659.
ismdreik ‘ to count '.
ispel ■ whistling v. p. 659.
tplurik ‘ to flee, spring ’ v. p. 664.
amnk, aurlk ‘ to ride '. Cf. Av. hnr-, Yaghn. voridki ‘ to ride ’.
iwis ‘ bridle ’, v. p. 664.
u'xdr ‘ ladder v. p. 66.3.
a'xrdnit ‘ stable ', < *dxtir-8dn- ?
'dyukun ‘ egg < *dGc)yak, cf. Prs. xdya, etc. But -nn ?
av'zdu ‘ heifer ', v. p. 664.
bo'yuzu ‘ frog ’. Cf. Prs. hazay, va°, etc., Av. vazaya-. With trans-
position, cf. Munji irityZ3ya.
bronsk, bran ‘ meadow ', Kal. bninz. Pal. brhunzv. Dam. brds. Po.ss.
< *mranz < Ir. *marza-. But Av. mardza- ‘ border country '.
bosk ‘ thick ’. Cf Wkh. bd), 8gl. vxzdk. etc. ?
6a.soy, hfls ‘ stream divided into many rivulets ’. Yd. (ydxyo) bnxsiyo id.
hd'than ‘ country v. p. 658.
biz'bdr ‘ eagle ’, v. p. 663.
lEAXIAIs EI,EMENTS IN KHOWAR
667
*bozi]i ‘ to play v. p. 660.
copik ‘ to gather, pick ' ; Wkh. ciip-, Sar. cev-, Yd. cur-,
cdrax ‘ spark v. p. 664.
coik, cJioik ‘ to be frozen ’. Pres. *c(h)dvm<in. Proh. from Prs. cd{h)ldan
‘ to feel cold ’. cf. Munji cly-.
ci'dhm ‘ brass pot v. p. 665.
cdrwelu ' headman of several villages carhu ’ village headman ' :
Prs. carhu ‘ fatness ’ (cf. Oss. bdsti-soi ‘ the fat of the land, a
prominent man ’).
chat ‘ lake ', v. p. 660.
ca'tir ‘ tent v. p. 660.
caxur ‘ spinning wheel v. p. 660.
do'yur ‘ finger nail v. p. 662.
dumik ‘ to card wool v. p. 664.
dreik ‘ to pour out etc., v. p. 663.
droxvm ‘ silver v. p. 660.
drazeik ' to load up ’ (> Kal. of Urtsun drdz-). Ir. darz- ‘ to wrap
up, to load cf. Sar. berz- ‘ to load ’ and Et. Voc. Psht. s.v. lehJ.
dmvarth ‘ door v. p. 660.
frosk ‘ right, true ’. Ir. I
fros ‘ muzzle ’, v. p. 663.
friieik ‘ to sniff ’. Cf. Wkh. ferx- ‘ to sneeze ' ?
gu'ddz ‘ trouble ’, v. p. 659.
ginvaldy ‘ peach ', v, p. 660.
gurzen ‘ garden ’, v. p. 658.
yeik ‘ to spin v. p. 663.
yacu ‘ leather ’. Ir. ? But cf. Gawar-Bati guci. Dam. gdci. Waig. gudi,
Kal. (Leitner) gao.
ydl ‘ polo ■ (Kal. gdl). Prs. ydl ‘ rolling ydluk ' pla\fing ball ’. Dialect
form < *gart- ?
ydleik ‘ to curl '. Prs. yuldla ‘ curls but no corresponding verb.
yalwanj ‘ kite ', v. p. 659.
ybli ‘ hare v. p. 6)65.
ydn ‘ wooden trough ’, Ir. ?
yon ‘ like ', v. p. 660.
yarik ‘ to turn (into) ', v. p. 662.
ydri ‘ upland pasture ’. Av. gairi- ‘ mountain ', etc. ?
yruc ‘ a bimch of grapes ’. Prob. lA. in spite of the y-. Cf. Gawar-
Bati urus, Pal. gruc. Dam. gii'rus, g'^ruc. Psht. of Dir gurur
‘ strawberry Kal. rue ‘ bunch of grapes ’ < Kho.
668
G. MOKGENSTIERNE —
yardik ‘ to turn round v. p. 664.
yrdnu ‘ pregnant v. p. 660.
yer'ndnu ‘ eddy v. p. 664.
yuzdr ‘ footsteps v. p. 659.
h^'dam ‘ limbs, body v. p. 663.
ha'ren ‘ mirror v. p. 660.
hmsk ‘ straight, true ’ (Dam. 3ska ‘ truth ') < Ir. *urzk, cf. Yd. urzuy,
etc. Kal. 'ujuk is lA.
ke'lec ‘ fathom ’, v. p. 658.
kren] ‘ wrinkle ’, v. p. 659.
kar'vas ‘ cotton ’ (Kal. kra'vas). Borrowed from Prs. karvds, °wds
(in its turn from Ind.), and again borrowed from Klio. into Yd.
kusk{omul) ‘ armpit ’. Cf. Prs. kas ?
Idsijd ‘ magpie ’, v. p. 662.
Idtep ‘ book ’, v. p. 658.
ko'vor ‘ pigeon Prob. lA., cf. Kal. ko'hdl- (Kal. u. ko'hor). But Waig.
kupto, Gawar-Bati kdpota are perhaps early Ir. loanwords.
khdm ‘ palate ’, v. p. 658.
khcna'gw ‘ sword ’ ; Prs. axangal, Sogd. xnyr, Yd. xu'gor, Ar. mnjar
(< Prs.).
last ‘ plain, steppe ', Kal. llmst. Pal. l‘‘kdst ‘ plain, flat From an
Ir. I- dialect, cf. Prs. dast ?
me'nu ‘ guest ’. < *memn-, cf. Prs. mihrndn, etc. ?
muryuluni ‘ down of birds ’, ‘ wool of ibex ’. Wkh. ntargilam ' goat s
wool ’.
nimez ‘ prayer ’, v. p. 660.
nask ‘ beak ’, v. p. 66-3.
rid'u'es ‘ grandchild ’, v. p. 658.
nivesik ‘ to write ’, v. p. 659.
naxoi ‘ pea ', v. p. 659.
pdlvrn (Kal. p>dl'im) ‘ soft, smooth ’. Wkh. pal(a)m ‘ soft, smooth,
fine powder Yd. polxm ' soft, smooth, fine ’, Sgl. pdhm. Also
Prs. palm ‘ dust ’ from an I- dialect ?
pe'lfsk ‘ rug ’, v. p. 658. Cf. Shughni pa'lds.
pidum/ustu ‘ finger-ring ’. Wkh. plougoM, but Yd. pargusce ( < *para»-
guMT). Not found in Prs.
posp ‘ wool ’. Probably an early loanword from Ir. pasni. Found in
most Dardic dialects ; Kal. pits, Bashk. pam, Gawar-Bati porn, etc.
pot ' polo ball ’, V, p. 662.
paxturi ‘ sunny side v. p. 660.
IRANIAN ELEMENTS IN KHOWAR
669
pdz ‘ breast v. p. 662.
pezemik ‘ to entrust v. p. 663.
phi' ran ‘ shirt v. p. 658.
phirwdl ‘ trousers V. p. 658.
phost ‘ skin v. p. 658.
roci ‘ fast v. p. 660.
ri'gis ‘ beard v. p. 662.
rum ‘ tail ’. Acc. to Turner (Nep. Diet., s.v. dumsi) early loanword
from Prs. In that case the only Kho. word with r- < Ir. S-.
The long u would be difficult to explain. Probably lA. = Skr.
luma-.
ro'men ‘ aspen v. p. 664.
rmzik, ranzuman ‘ to shiver ’ (Kal. ronz-. Dam. rdz-). Probably Ir.
ranz- < *ra-rz-, cf. Prs. larzldan, Psht. rezdeddl, etc. Welsh
Gypsy rizer- < *rirz- ?
rost ‘ daylight ’. IVkh. raM ‘ day Sgl. rast ‘ dawn ’.
riwis ‘ rhubarb v. p. 659.
roxeik ‘ to leave behind ’. Ir. ?
roxm ‘ embers v. p. 662.
sa'puk ‘hoof’, v. p. 662.
sor ‘ head ’. The o indicates a rather early borrowing. Yd. s&ro ‘ ear
of corn ’ from Kho.
sareik ‘ to appear v. p. 662.
srung ‘ horn ’. Cross between *sring (< srnga-) and Av. sru- ? The
modern Pamir dialects all have s < sr, for instance Yd. su.
-saxa, V. p. 664.
sauzeik ‘ to sew, prepare ’. Ir. ?
sa'yurj ‘ hawk ’, v. p. 664.
sunj ‘ needle ’, v. p. 663.
sinjur ‘ jujube-tree '. Prs. sinjid Yd. s9ztyo. Munji sijia. Anc. loan-
word from Ir. (with s-f > s-j) or vice versa 1
sexe (O’Brien) ‘ stumbling ’. Cf. Prs. saxidan, Yd. .s“x«y- ‘ to
stmnble ’.
sexik ‘ horn ’, v. p. 658.
soxeik, saxcumdn, soxs- (O’Brien, Ganj-i-Pukhto) ‘ to pass, cross a
pass, to forgive ’, in tbe G.-i.-P. ‘ to pass ’ (about the time).
Kal. saxsim from Kho., Sgl. sxxs-, Wkh. ioxs- ‘ to pass ’. Probably
< *saxsa- from Ir. sak- ‘ to pass (the time) ’.
sd ■ black ’. Wkh. Slu Sgl. sui. Ir. ?
sabllki ‘ lucerne ’, v. p. 658.
670
G. MORGENSTIEBNE —
sangur ‘ guts v. p. 663.
sonthn ‘ raven v. p. 662.
sa'pik ■ bread ’. AATch. xapik, Shgh. sapik, etc., v. Rep. Lingu. Miss.
Afgh., p. 77.
tanau ‘ rope v. p. 664.
to))k ‘ .shallow ' ; Prs. tiimik. etc. Also Kho. tonuk ‘ thin, delicate etc.
ta' par-zlng ‘ battle-axe v. p. 660.
trus'ni ‘ thirsty v. p. 664.
trispoy ‘ dandelion v. p. 664.
tu'vek ‘ gun v. p. 659.
toxmiran ‘ daughter-in-law ''(??). v. p. 661.
wd ‘ again, Prs. hdz '. Prs. wd is used only as a verbal prefix.
tve- ' without ' < Ir. *(a)^e-, Prs. be. ive-iatu ‘ fatherless ' ; we-xdl
‘ tasteless ive-soru ‘ widow ’.
wdy ' price v. pp. 662, 664.
tvdru ‘ rain v. p. 662.
W'or ‘ smell ’ (> Pal. 5r). < */3o8, cf. Wkh. vuh Werchikw. loan-
word veri.
werkhd ‘ lamb ’ ; icerkhdlu ‘ram, three years old ’. Scarcely borrowed
from Ir., AVkh. icurk, etc.
warum ‘ felt, numda ' ; tvardnk ‘ posteen ‘ ; sa tvarioii ‘ siah-pos ’. Ir. ?
wrenjeik ‘ to fry v. p. 664.
wdrz ‘ above '. v. p. 664.
tvdrzidini ‘ bolster v. p. 663.
wrazni ‘ pillow v. p. 663.
wrazun ‘ wing ’ (Kal. bazur-gundi) ; uardznen lo (Kal. torazan-lou)
‘ flying squirrel ’ ; poshvrdzun, poz'wdzur, postw° (Kal. pozurazan)
■ bat ’. Cf. Yd. vdrzeyo, Munji vozsrgo, Psht. wazar, etc.
tve-soru ' widow ’ (> Yd. was9rwo). Ir. *^e-sar ‘ headless cf. Kal.
a-sisa ‘ widow, he-sar
wezen ‘ evening ’ < *uz-agana-, cf. Av. uzayara-, uzayeirina-. Yd.
iziane.
'xacvw ‘ owner ’. Ir. ? ?
xdl (Kal. khaV, Yd. xdl from Kho.) ‘taste’. < Ir. *x'"dl <xwdS,
Prs. xiu'dj.
xel sweat , v. p. 662 (Pash, of Kurangal ke'le from Psht.).
xmnik, axo/nik, xu'amik ‘ to descend ’. AVkh. xam-, cf. Sar. xdvs-,
Yd. xafs-, Prs. xambdmdan ‘ to bend ’, etc.
xer ‘ stench ’. Ir. ?
xora ‘ mill ’. Cf. Sgl. xuBdri. Yd. xiryo, etc.
lEAISriAN ELEMENTS IN KHOWAR
671
ntr ‘ other h Ir. ?
xu'ri ‘ wild rose v. p. 663.
xeseik ‘ to entreat Ir. ?
Task ‘ soft ’. Ir. ?
xosp ‘ dream '. Prs. xusp ‘ asleep '. But s ?
xatan ‘ house v. p. 660.
xausik, xmvsik ' to request ’. Ir., cf. Arm. xawsim ‘ I say ' 1 ?
o'u'i-dn ‘ happy ’ (Kal. x/son, Dam. kho'san, Bashgali knsan from Kho.).
Cf. Sogd. °xivasant (yios'nt) ? Or Prs. xusiidii ?
To‘//d«M ‘ belly '. Ir. 1
xaz'gum ‘ carrot Kati (Chitral) gaz'gfon. *gaj{<i)k{ -y- um) of lA. origin,
but perhaps influenced in the ending by Sgh, Prs. mlyam
‘ turnip ’.
yormdn ' woe, arm-dn ' (in poetic formula). Early loanword from Prs.
ermdn wdth Kho. e- > go- (cf. yo'mun ' winter ').
yuran Bidd.. yurdn O'Brien ‘year*, yerwdn Bidd. ‘solstice .
Connection with Av. ydr- ' year ’ is very improbable.
yozun{i() ‘ yoke '. < Middle Ir. *yozan = Skr. yojatia- ?
zoik ‘ to agree to '. Ir. ?
zay ‘ fat, grease '. Yazghulami zey ' liquid fat Psht. ydz ' fat of the
kidneys ’ (and Oss. qa{n)z ‘ marrow ' ?).
zo'meik ‘ to yawn ', v. p. 662.
zdm ‘ old ’. Prs. zar ' old man Yd. zor.
zero ‘ yellow '. Ir. *zcrt + c, cf. Seistani zerca or zanh ' rust disease
of wheat '. Shgh. zerec ' rust ’. Kho. zerbali ' yellow rose < *zerd-
iraU 1
zar ' poison ', v. p. 660.
-zeri ‘ young animal v. p. 660.
zirymk ‘ to lament ', v. p. 664. ‘
^ Also the other Itardie languages contain some early Ir. loanwords. 1 ha%e
mentioned Shina, etc., yuspiir ' prince Rep., p. 76. Others are c.g. Kal. parte (Pam.
pareci) ‘ sieve ' : Parachi parjeon ; Kal. sn*itu^ ^elas ‘ small fiat basket , -^ohola
‘ big, deep basket ’ < *tinpat- : Prs. aapad. etc. ; zhantyak ’ child-beanng : Av.
^qdya - ; Gawar-Bari ziitak. Bashkarik jatak ‘ boy(s) ’ : Phi. zaink ; Pameli rOpak
‘fox’ (or lA ?) ; Ashkun butXml ‘almond’: Prs. hadain, etc.; Palola hetmgala
' tomato, egg-plant ', Kohistani 8hina bhilmyni ■ Prs. Unhmjhu < lA. (through Ir. ?).
The History of p in Kanarese
By A. N. Narasimhia
T he change of p, initial and intervocalic, to h in Kanarese was
noticed by Western scholars early. Caldwell ^ says “ the initial
p of nearly aU the words — whether they are pure Dravidian or Sanskrit
derivatives — -ehangcs to A A more careful study of this change of
p to h was made by Kittel in his Grammar.^ He points out that initial
p was changed to h in Kesiraja’s time (1260) optionally. But he has
not given evidence from the inscriptions to show when p changed
to h.
Below are summarized the results of an analysis of the inscriptional
forms given in the appendix to this article.
Initial p-
Initial p- was maintained till the end of the ninth century. In the
sixth and seventh centuries the following words appear : —
poll (=NK. hali), parvarumdn (hdruvarannu), palum (hdlu).
pdvu (hdvu), pin (himde), piriguni (hiri), pulla (hullannu), pulu {kulu),
puni (hdnu), peran (hera, hora), perjediya (hejjede), pelcugc (heccu),
periyd (hiriya), pelda (helida), pokka (hokku), pogevogi (hogakogi),
pomgol- (homgol-), podeda Qiodeda), poragu {horagu).
So, too, in the eighth and ninth centuries (see Appendix).
Forms with h- <. p- appear first in the tenth century. The h-
forms are more common in prose, especially where the boundaries of
lands granted to mathas and temples are described ; in verse very few
h- forms occur.
' CDG., pp. 156 and 157.
[The following abbreviations are used in this article :
CDG. = K. Caldwell ; Comparalire Grammar of the Dravidian Languages.
EC. — Epigrapkia Camatica.
El. = Epigrapkia Indica.
KBB. = Karnataka Bhdsa Bhusana.
KG. ^ r. Kittel : Grammar of the Kanarese Language.
KSS. = Karnataka Sabdanusasana.
KVV. = Karnataka Kavyavalokana.
MK. = Middle or Mediaeval Kannada.
NK. = New (Modern) Kannada.
OK. = Old Kannada.
S. = Sutra.]
= KG.. Sections 64, 184, 223, 224, 370.
674
A. N. XAEASIMHIA
In the eleventh to the fourteenth centuries h- forms are more
frequent than 'p- even in verse ; they are even more numerous in prose.
In the fifteenth and sixteenth centuries v'ery few forms with p-
are found ; and in the seventeenth century p- occurs only in verse.
Already in the thirteenth century this h- <p- was itself beginning
to disappear: in a.d. 1219 we find ogu {—hogu), odar {= hodar),
El., V, p. 261.
To-day in the colloquial speech of the uneducated, h- < p- has
practically disappeared before vowels of all qualities. The uneducated,
trying to speak well, often insert an h- when it is not etymologically
required.
Intekvocalic -p-
Primitive Kanarese possessed no -p-, only -pp-. Dravidian *-p-
had already become -v-. This is seen even where -p- was the initial
of the second member of a compound as early as the seventh century :
edevidiyal, Edevolalnadu, pogevogi, muvelmura, marhjuvol, bittavdl,
salvavol.
This -pp- was, like other double consonants, shortened to -p-
after a long vowel, or in the terminational element of polysyllabic
words or even in dissyllabic forms of certain auxiliary verbs and
pronouns.
In the seventh century the causative suffix appears with -p-
{< -pp-) in kalupe, and as an auxiliary verb apar appears beside
dppar.
This -p-, like initial p-, became h in the tenth century ; hohan
(a.d. 910), After a short vowel in a termination ; nmlahidar (a.d. 1004).
The change affects -p- of Sanskrit loanwords ; nrham (a.d. 1382)
< nrpam.
The inscriptions have the following examples of -li- < -p- < -pp--.
1004 nadahidar sattarendu, EC, i, 46 ; 1172 ulihim, EL xv, Madagihal ;
1175 alihidade, EC. i, 65 ; 1182 hoha (48), El. xiv. Kurugod ; 1218
hoharu appa, EC. iv', Hg. 23 ; 1223 hoharu (6), EC. vii, Sk. 175 ;
1229 aluhidava (12), EC. iv, Gu. 19 ; 1282 appa (19), EC. ii, 334 ;
1295 kondamtaha (11), EC. iv, Yl. 44 ; 1300 baharu (37), EC. iii,
TN. 98 ; ilihikottu (27), baha (35) ; 1317 alihida (26), hoharu (30),
EC. iv, Ch. 116 ; 1368 hoharu (31). baheu (25), uljamtaha (19), EC.
ii, 344 ; bahevu (16), EC. iii. Nj. 117 ; 1370 iharu (15), hoharu
(16), EC. iv, Ch. 97 ; 1376 hohaga (8), EC. vii, Sk. 57 ; 1390
hoharu, EC. i, 39 ; 1391 hoharu. EC. iv. Yd. 1 ; 1932 appudakke
THE HISTORY OF p IN KYNARESE
675
(21). EC. vii, Sh. 11 ; 1406 alupidavam, EC. iii. Sr. 105 ; 1409 bahevu
(79), antaha. EC. ii, 255 ; aha, EC. V'ii. Sh. 70 ; 1431 hahiri (53).
EC. vii. Sh. 71 : 1437 hoharu. EC. iii, ill. 4 ; 1444 hOhanu (11. 12. 13),
EC. iv, Yd. 7 ; 1477 hoharu (28), hahiri (20), EC. iii, ild. 77 ; 1484
hoharu (18), baheu (21). EC. iv, Xg. 59 ; 1500 hOharu. EC. ii, 395 ;
1500 hoguvaru, EC. ii, 340 ; 1517 hoharu (10. 11), EC. iii. ily. 5 ;
1539 bahevu (7). EC. ii. 225 ; bahenu (13). EC. ii, 224 ; 1544 appa
(12), hovaru (35), EC. i. 10 ; 1550 bahadu (15), hoharu (19), EC. iii,
My. 50 ; 1557 boha (38). EC. vii, Hg. 9 ; 1564 kaluhi (8), yihan (14),
EC. iv, Yl. 29 ; 1576 bahiri, EC. iv. Yd. 59 ; 1620 alupidavanu (13),
EC. iii. My. 17 ; 1645 yiha, EC. iv, Ch. 124 ; 1663 bahiri (53), EC. iii.
My. 13 ; 1670 bahiri (20), EC. iv, 1 and 9, 119 ; 1672 bahudu, EC.
vii, Sk. 215 ; 1678 baha (25). EC. iii, Sr. 94 ; 1593 Yihadu, EC. i, 11 ;
1753 yiha, EC. iv, Ch. 128.
Further, the pronominal forms of OK. antappa, intappa. entappa
have become antaha, intaha. entaha in MK. ; in NK. anthn. inthd, entha.
Survival of p- Forms
Many words are still used both in hterary composition and
colloquial speech with an initial p-. A few of them are loanwords :
pakka ‘ a side ', pagadi ‘ tribute, tax ’, paccadi ' a kind of pickle
patti ‘ a squared rafter ’, pattu ‘ hold, seizure patte ’ the rind or bark
of trees ', paduvalu ‘ the west ’, padde ‘ maturity '. palaka ‘ practice,
habit ’, padu ‘ experiencing papa ‘ a small child '. pdiu ‘ a share ’,
pedasu ‘ hardness ’, poUare ' a hole in the trunk of a tree ’, pokari ‘ a
profligate ’, poti ‘ joining ', pdnisu ' to string together ’, polu ‘ wasting ’.
In some cases p- and h- forms are both found in use, but
differentiated in meaning : —
A. From the same Root
pdlu s.n. • ruined place ' used in proverbs which retain archaic
forms ; hdhi vb. ’ to be spoiled, to ruin
pettu s.n. ‘ a blow ' ; vb. ' to beat as with a hammer ' ; hettu vb.
‘ to thrust (with the penis) '. perhaps to avoid the use of the expression
which had acquired a vulgar meaning, pettu was borrowed from the
hterary language in the meaning of ‘ a blow ’ and " to beat ’.
B. From different Roots
hdl ‘ milk ’, pdl ■ a share ' ; hadn ' to sing ' and ‘ a song ', padu
‘ difficulty experienced ' ; hurudii ' rivalry ’, puriidu the pollution
after the birth of a child ’ ; holu ‘ to resemble ', pdlu ' waste
676
A. N. NARASIMHIA
Evidence from Literature
Grammars : There are four grammars of the ancient Kanarese
language. But in a book on rhetoric, a.d. 877, Kavirajamdrga, the
portion relating to grammar is comparatively small. In the part on
rhetoric we find no h- initial or otherwise in the pure Kanarese word.
I, 1045. Karnatakakavyavaldkana is in verse. There is no h- at all.
II, 1045. Karndtakabhdsabhusana by Nagavarmma includes
A- in the alphabet (KBB. 6). In Sutra 115, he says “ pe hova ” —
that p > A often initially and intervocalically, as the examples he
gives prove ; pahge > halage ; kdpu > kdhu.
III, 1260. Sabdamanidarpana by Kesiraja. He includes A- in
his account of the alphabet (SMD., S. 33). In S. 159, he states that
single p of pure Kanarese words often changes to A and that such a
change is beautiful (sundaram), but he definitely points out that long
p {-pp-) does not become -A- and says that it is dustara or difficult.
Examples given are upparam, kappuram, tappu, bippandam, niuppu,
muppuri, soppu, heppu. But in S. 150 -pp- of the pronominal antappa,
intappa, entappa are changed into a single -A-, i.e. antaha, intaha,
and entaha.
Further, in sutras 254, 258, and 292 he points out how Sanskrit
loanwords with p- change to A- in Kanarese ; pisunam > hisuimm,
pdsa > hdsa.
The disappearance of A- im’tially is not unknown to him. In S.
271 he gives hirhgu > ingr(, hamsapinda > ancevindu ; harhsa > amce,
himgulikam > imgulikam.
Use of unnecessary A- was known to Kesiraja (Smd. 269) : agni
> ^aggi-
IV, 1604. Karndtakasabddnusdsana by Bhattakalariika.
In S. 6 he includes A in the Kanarese alphabet : in S. 145 p > h
often, e.g. pattike > hatlige ; pJvara > hivara, gopura > gokura.
In S. 296, Tihurdntakan and Gohuram with -A- are referred to as being
correct. In S. 497, appam > aham, appudu > ahvdu. In S. 576 he
states that Am (< pu) is to be affixed to hd ‘ to come ’ to form verbal
nouns.
Appendix
Eighth Century . — 726 punname, EC. iii, Tn. 1 ; 740 puttade (21),
EC. iii, My. 55 ; poppandu (16), EC. iii, My. 55 ; 750 pin, EC. ii, 79 ;
pannirwarparwar, pattu, EC. iv, 149 ; 4 pervaUame, EC. iv, Hg. 4 ;
THE HISTORY OF p IN KANAEESE
677
776 paleya (66), EC. Ng. 85 ; 800 poldu, pogi, periya ; ii, 35, 4 ;
800 perbala, EC. iv, Sr. 160.
Ninth Century. — 810 pusuvan (7), EC. iii, Nj. 26 ; 830 puttida (5),
EC. vii, SK. 283 ; 870 palaram (9), EC. iii, Nj. 75 ; 870parvvammam
(10), EC. iii, Nj. 76 ; 884 padinayduvarisadandu, EC. ii, 394 ; 888
padinentaneya (5), EC. i, 2 ; peddoregareya (6), ponnum (14) ; 890
pattugadyanada (8), EC. i, 3, 8 ; 890 pervvayala (71), peronie (73),
pallame (74) ; perolve (65), pervaltiya (75), perggolliya (76), EC. iv.
Yd. 80 ; 898 Elecaga palliya (3), EC. iii, Nj. 98.
Tenth Century. — 900 puttegu (6), EC. iii, TM. 115 ; 907 Polalasetti
(6) , padinaydupananum (11), pattondiya (8), EC. iii, ND. 14 ; 910
lokake hohan, EC. iii, Sr. 134 ; 930 perggedegalum (6), paridava (14),
EC. iv, 149, 116 ; 950 pandiyum (11), piriya (12), EC. iii, Md. 41 ;
972 punnameyum (7), piriya Holma (5), EC. iii, Nj. 183 ; 978 piridu
(18), Perggadururii (11), Peddoregare (13), EC. i, 4 ; 960 puttidom (15),
puttidar (6), EC. iii, Tn. 69 ; 982 perarorbbarum (41), paruvalli (61),
podisuva (6), pogalisu (3), EC. ii, 134 ; 982 pogalisal (113), peldapem
(31), porage (80), pusivude (45), piridiva, EC. ii, 133 ; 985 perggade
(2), poge, El. xvii, p. 170 ; 995 pogale, EC. ii, 121 ; 1000 punpame
(7) , Bihagamup(Jana (9), Hancadarmasetti (11), Bokaballi (13),
Hakadivadi (14), Gorahajli (15), EC. i, 5.
Eleventh Century. — 1004 nadahidaresattarendu (6), EC. i, 46 ;
1007 ponnol (14), penadotti, palaram (22), EC. iii, Tn. 44 ; 1012
eraduhottina (30), payisidam (11), EC. iii, Sr. 140 ; 1019 parbbi (14),
halladi (29, 30), holakke (29), holada (31), budomta (32), pudorhta
(40), puduke (16), EC. vii, Sk. 125 ; 1021 padineydu kolaga (18),
punusevaram (20), pumbolanurh (21), EC. iv, HG. 16 ; 1021, 1036
badinentu (22), pudornta (23), EC. vii, Sk. 126 ; 1049 Hosanada (28).
ponnarakottu (31), EC. iv, Gu. 93 ; 1050 panneradam (9), EC. i, 30 :
1050, 1057 pervvayal (18), polipare (16). pogi (26), pogale negalutuin
(27), EC. iv, 149, 18 ; 1058 Haruvanahalli, ArakanahalU (12), Hennel-
lunabhumi (21), kariigonda hallihamgalabani (26), hunise (29), EC. i,
35 ; 1060 Huliyakereya (11), Hegga naleya (14), haduvanakolada
(15), Heggerege '(19), halliya (21). hiriya (23), EC. vii, Sh. 6 ; 1063
parekara (24), harekara (25), hiriyakereya (27), horavarige (24),
EC. vii, Ci. 18 ; 1070 Mudanahaduvana (6), Hosavalli (8), EC. i, 49 ;
1070 Hosavallisiddhesv'ara mahadevargge, posavalli (6). EC. i, 50 ;
1071 heggade (57), besar (94), El. xv, p. 337 ; 1075 hajla (46—51),
hadada (52), hanneradu (54) ; 1 hadinaru (55). halla (55), pudornta
(53, 55), El. XV, p. 96e ; 1076 perggade (33), paduvarggam (24),
678
A. X. XAKASniHIA —
paimeradu (30) ; Kotehalasavanta (35), hesara (38), EC. vii, HE. 14 :
1079 baraha (49), posavolala (33), EC. iv, Ng. 56 ; perggadegala
(29) ; 1085 poltixkkum (48), peklu (59), poydam (59). liarige (47),
barige (51), Hennavurada (51), pere (44), EC. vii, Sb. 10 ; 1087 boda
(24), EC. iv, Yd. 2 ; 1089 Hosavura (3), posavura (4), perottimge (5).
palaram (5), Hosavura (6), EC. vii, Sk. 298 ; 1095 basuvum baruvanam
konda (58), EC. i, 57 ; Buvanahalliyam (9). EC. iv, Kp. 49 ; 1097
horege (55), herimge (55), bonnam (56), pana, paduvalu, El. xvu,
p. 182 ; 1099 Panasogeuivasi. EC. iv, Yd. 24 ; 1100 pesarwadedar
(23), EC. ii, 69.
Tivelfth Century. — 1104 putti (20), hermmagal (39), EC. \di, Sk.
131 ; 1107 Hosavolal (4), Posavolala (5). Sogepalli (5), EC. iv, HG.
107 ; 1107 herimge (14), honna (15), her (17), Panumgall (9), panav
(15), perin (17), El. xiii, p. 12 ; 1110 battu (33), and all the rest begin
with p-, EL XV, 26 (Mutgi inscription) ; 1112 paduvalu (80), pom,
ponna (83), pana (83), boda (77), bola (77), balla (78), battu (78),
EL xiii, p. 36 (Ittagi) ; 1113 pogalvudujanam (26), EC. ii, 126 ;
1115 perggade (44), Hoyasalamaharaja (156), EC. ii, 127 ; 1118
harubageyam, EC. ii, 125 ; 1120 hesarittu (11), halli (11), Hoysala (5),
EC. iii, 32. 43 ; 1123 hadimuru kamcina ho]avigeya (54), polvavarar
(17), Poysaladev'ara piriyarasi (42), EC. ii, 132 : 1124 Herm-
madigavundana (3), EC. iii, Nj. 194 ; 1125 Halasige and palasige,
EL xiii, p. 298 ; 1135 huvinatotamum, EC. ii, 384 ; 1144 hamdiyan
(6), EC. iii, Md. 22 ; 1145 p6 p6 vadi pogendu (148), EC. ii, 140 :
1147 bore, hirngidudu, bal, bom, pannirpuligere, padedu, hattar,
bola, hala, horeya, balla, El. xvi, p. 44 ; 1148 honnalakottu (18),
baraha (2). EC. iii, YJ. 110; 1150 hanavinalekka (20), hermmagal
(39), EC. vii. Sk. 131 ; 1152 Tailaha, Hanumgall, panurngall, El.
xvi, p. 36 ; 1159 haduvalu (62). huniseya (62), biriyaru (62), EC. ii,
345 ; 1162 halgalam (31), hoheyam (32), bun (33), and the rest are
allp-, El. xviii, p. 212 ; 1163 Huljarajam (108), Heggadekorayyanum
(103), EC. ii, 64 ; 1172 Hemmadi (12), Ldihim (32), hiridurn (82, 94)
(in verse), both p and h in prose, EL xv, IMadagihal.; 1175 Hoysana
(4), Biluhunadu (7), alihidade (16), hoharu (19), EC. i, 65 ; 1175
Huvinapadage (3), himdegade (3). honnabeddaleya (4), EC. ii, 242 ;
1179 hodabatteya (43), hulumadiya (43), Hiriyadevarabettakkam (44),
haduvalu, hiriya, haj|a (44), EC. ii, 397 ; Hoysalana (32), Torabanam
(33), Torapa (63), hiriya (67, 71, 73), bola (73), paduva (71), El. xix,
p. 226 ; 1180 poy. pesaran (4), Hanariigallu (17), Hoysalavlra-
ballala (19), Kaluballi (22), EC. iii, Tn. 106 ; 1182 piriya (36), paduval
THE HISTORY OF p IN KANARESE
679
(40, 47, 48), pu (40), Mhe (40), hiriya (44), hoha (48), El. xiv
(Kurugod) ; 1184 pasuge, periiiige, hala, hasuge. hana, El. xvii,
p. 189 ; 1195 puttidam (7), pididam (16),. piriigugum (32), hanamuru
(56), EC. ii, 335 ; 1199 hattigalegadol (14), EC. iv, Ng. 47 ; perar ;
1200 hogalu (11). halli (11), hiriide (11), EC. iv. Kp. 47.
Thirteenth Century. — 1203 bore (46), hasumbe (47), Hotteyya (43),
perggere (41), hola (48). huttida (48), EC. vii, Sh. 88 ; 1204 peridodam
(52), perim (54), otherwise all h-, El. xiii, p. 16 ; 1206 hana (16),
hiriya (13), hokkade (28), EC. ii, 333 : 1210 Huligere (13), hitu (13),
paduva, padeda (40), El. xix, 194 ; 1213 Harahondanamaga (3),
Kondahohali (4), EC. iii. Ml. 37 ; 1214 perggade (12), heggade (14),
EC. vii, Sk. 243 ; 1217 haduvanahuniseya. EC. ii. 170 ; 1218 rahuta,
hoharu, appa (20), EC. iv, Hg. 23 ; 1218 hesara (49). Hadavalagoggi
(21), pusi (9), EC. vii, Sh. 5 ; 1223 hadinaraneya (2). halli (4), hohaga
(6), halaram (6), EC. vii, Sk. 175 ; 1229 hodaru (13), Heriyanada (8),
aluhidava (12), EC. iv. Gu. 19 ; 1235 hesarithdarh (37), hoda (41),
pesarvetta, pokkade (17), EC. iii. Md. 121 ; 1246 Hiriyabettadi (2),
EC. ii. 165 ; 1255 hogi huttida, EC. i. 6 ; Hoysana. appa, aha ; 1276
samauvitavaha (for appa) (13), EC. iii, Md. 70 ; Sthavaravaha (15),
EC. iii, Md. 70 ; 1282 horagiigi (26). appa (19). EC. ii. 334 ; 1285
baluhirhde (12). Hovisala (8). Haradayva (26). EC. iii, Md. 62 ; 1290
haUada (20), herobbe (22), EC. iii, Tn. 27 : 1295 Homma (11),
Kondaihtaha (11), EC. iv. Yl. 44 (for appa) ; 1300 baharu (37), EC.
iii, Tn. 98 ; ijihikottu (27). baha (35) (for barppa). honnonu. EC. iii.
Tn. 98.
Fourteenth Ceritimj. — 1317 alihida (26), hoharu (30), EC. iv, Ch.
116 ; 1325 hattu (10), Hosahalh (10), hadinaidu (22) ; 1348
Hanagalliriige (4), EC. i, 63 ; 1360 hogi (3). Hulukoda Chikkannayya
(3), EC. i, 67 ; 1368 hoharu (31). honniihge (25). baheu (20), ullariitaha
(19) (for appa), hadinentu (13), EC. ii. 344 ; bahe\u (26). EC. iii,
iSTj. 117 ; 1370 iharu (15). hoharu (16). EC. iv, Ch. 97 ; 1376 hayidii
(8), hohaga (8), EC. vii. Sk. 57 ; 1377 biihnahani (20). EC. vii, Sk. 35,
for vijnapana ; 1382 nrpam and nrham (king), pogalgu (14), EC. iii,
Nj. 21 ; 1382, 1390 halla (7), halli (8), hariva (8), pesarainttu (23),
hunisedalu (25), hoharu, haiiidiya (31). EC. i, 39 ; 1391 hottina (6).
hoharu, EC. iv. Yd. 1 ; 1392 excepting appudakkc, every p- and -p-
is h- and -h-, EC. iii, Ml. 47 ; 1400 ahudu (21), EC. vii, Sh. 11.
Fifteenth Century.— \m halH (7). alupidavani (13), EC. iii. Sr.
105 ; 1409 bahevu (79), hoiiina (78), asadharanavamtaha. EC. li. 253 ■.
abhivrddhigal aha hage (27), EC. vii, Sh. /O ; 1431 anubha^isi, bahiri
680
THE HISTORY OF p IN KANARESE
(53), EC. vii, Sh. 71 ; 1437 hoharu, EC. iii, MI. 4 ; 1444 hohanu (11,
12, 13), EC. iv, Yd. 7 ; 1477 hoharu (26), bahiri (20), EC. iii, Md. 77 ;
1484 hoharu (18), baheu (21), EC. iv, Ng. 59 ; 1500 hoharu, EC. ii,
395 ; 1500 hoguvaru, EC. ii, 340 ; all h- and -h- except in verse.
Sixteenth Ceiitury. — 1509 sahodararaha (10), EC. ii, 228 ; 1513
hage (25), EC. hi, Gu. 3 ; 1517 hoharu (10 and 11), EC. iii. My. 5 ;
1539 bahevu (7), EC. h, 225 ; 1539 bahenu (13), EC. h, 224 ; 1539
adahagiralagi (7), adahanu (8), EC. ii, 224 ; 1544 homnu (28), Hanasoge
(27), halaru (33), Hosahalli (27), appa (12), Hosagadde (54), hovaru
(35), EC. i, 10 ; 1550 bahadu (15), hoharu (19), EC. hi, My. 50 ; 1557
yi hattanada (10) (for pattanada), hosa (38), EC. vh, HI. 9 ; 1564
bimnaha (8), kaluhi (8), yihari (14), EC. iv, Yl. 29 ; 1569 adahu (10),
EC. iv, Hg. 41 ; 1576 bahiri (30), EC. iv. Yd. 59.
Seventeenth Century. — 1620 alupidavan (13), EC. hi. My. 17 ;
1634 adahu (18), adavakotamtavaru (40), hoharu (49), EC. h, 352 ;
adahina (23), adava (24), EC. ii, 250 ; 1645 Haradanahalliyalu yiha,
EC. iv, Ch. 124 ; 1650 hakiyiruvadu, EC. iv. Yd. 40 ; 1654 Verse
portion; pasivamte (11), pesaran (15), posatu (14), irppudu (13),
peldan (35) ; Prose portion : huvadiganige (77), hola (90), Hosahal}i
(80), EC. iv, Yl. 1 ; 1663 bahari (53) and all h-, EC. iii. My. 13 ; 1670
bahiri (20), EC. iv, Hb. 119 ; 1672 undu bahudu, EC. vh, Sk. 213 ;
1673 nityotsaha (4), EC. h, 390 ; 1678 nadedubaha (25) and all h-,
EC. ih, Sr. 94 ; 1693 yihadu, EC. i, 11.
Eighteenth Century. — 1753 samudradalliyiha (5), EC. iv, Ch. 128 :
1775 pra-u-da pratapa, EC. iv, YI. 4 ; 1782 prahuda pratapa, EC.
i, 12, 13, 14 ; 1800 praudapratapa, EC. ih, Sr. 8, and all h- in pure
Kanarese words.
Grammairiens tardifs et dialectes du prakrit
By Luigi A Nitti
QIR GEORGE GRIERSON, en reprenant apres Lassen I'etude
^ de I’unique MS du Prakrtakaljxiiaru de Ramasarman Tarkavagisa
(No 1106 de la Bibliotheque de 1 India Office) et en le comparant au
PraJcrtasarvasva de Markandeya, a reconnu I’existence d'un groupe
de grammairiens prakrits, fideles a I’enseignement de Vararuci pour
ce qui concerne le prakrit principal, mais caracterises en outre par
le nombre beaucoup plus grand de dialectes qu'ils traitent. II les a
nommes “ grammairiens orientaux ” et s’est efforce pendant de longues
annees de divulguer leur doctrine et d’utiliser leurs donneesd
Mais I’interet eveille par ces textes n'a pas ete grand, les savants
ont montre une certaine hostiUte a I’egard des nouveaux venus.
D’apres eux I’essentiel de la grammaire prakrite se trouve en germe
dans Vararuci et est pleinement developpe par Hemacandra . I'un et
1 autre bien connus et magnifiquement edites. Markandeya et
Ramasarman ne peuvent pas etre anterieurs au XV® siccle et le
foisonnement des dialectes n'est que la preuve de leur epoque tardive.
Leur temoignage n’a done aucune valeur et ne sert qu’a cmbrouiller
da vantage la question deja si compliquee du prakrit et de ses
subdi\’isions.
Que Markandeya et Ramasarman soient recents, e’est un fait
indiscutable, mais commo remarque Grierson : A still later writer
than Markandeya was Professor Pischel, and the same opponent who
objects to the one will quote the other as infallible. It is not a question
of date but of sources and of critical acumen.” *
Le prakrit n'etait pas davantage parle au temps de Hemacandra
qu’a celui de Ramasarman et de Markandeya. On pent done admettre,
en raisonnant a priori, e’est %Tai, mais en se basant sur I’analogie
de nombreux cas dans I’histoire de I'lnde, que les grammairiens plus
recents soient pour ainsi dire archaisants et qu’ils nous transmettent
le reflet d’une doctrine tres ancienne, tandis que Hemacandra par
exemplo ait ete de son temps un novateur.
^ The Eastern School of Prakrit Grammarians and Paisaci Prakrit. Sir Asutosh
Mookerjee Silver .Jubilee Volumes, pp. 119—141 ; ‘‘The Prakrit \ ibhasils,” JPAS.,
1918, pp. 489-517 ; “ The Apabhramsa Stabiikas of Ram.a Sarman (Tarkavagisa),”
dA., li, 1922, pp. 13-23, and 1923, pp. 1-8, 187-191 ; “ The Pr.ikrit Dhatvadesas
according to the western and eastern schools of prakrit grammarian.s,” Memoirs
of the Asiatic Society of Benyal, vol. viii. No. 2, pp. 77-170 ; “ The Sauraseni and
Magadhi Stabaka.s of Rama-Sarinan (Tarkavagisa),” 7.4., Ivi, 1927, and Ivii, 1928,
etc., etc.
® The Eastern School of Prakrit Grammarians, etc., p. 122.
VOL. Vm. PARTS 2 AX I) 3.
44
682
L. KITTI
II est heureux que des faits viennent maintenant etayer cette
hypothese.
La theorie que I’anciennete d’un grammairien est inversenient
proportionnelle au nombre de dialectes qu'il traite, remonte, comme
la plupart des idees qui ont cours dans la philologie prakrite, a Ch.
Lassen : “ Ita ut — ainsi qu'il conclut apres un long raisonnement —
inter Prakriticae loquelae praeceptores is novissimus possit baberi,
qui maximo dialectorum numero locum in commentariis suis con-
cesserit.” ^ Pischel a accepte cette doctrine sans discussion et s'en
est meme servi pour essayer de dater Kramadisvara. Kramadlsvara
ne peut a son avis etre plus ancien que Hemacandra parce que il
“ nomme ” un plus grand nombre de dialectes.^
Or cette theorie — generalisation bardie mais comprehensible a
I’epoque de Lassen — n'est plus guere acceptable.
De son dernier voyage au Nepal, M. Sylvain Levi a rapporte
entre autres MSS, la photographic d'une grammaire prakrite ecrite
par Pirrusottama. J'ai ete chargee d’en assurer I’edition qui est
maintenant prete et qui sera, j'espere, bientot imprimee. Le dernier
colophon du MS nous donne une date et nous permet done de tenter
r identification de son autexm ; iti jnirusottamadevasya paisaeika-
siitram xamaptam \ sam 385 jepulikhitam ultamasrljhdnena saptativar-
mdhikavayasa.
L’an 385 de here nepalaise correspond au 1265 a.d.
Le grammairien et lexicographe Sanskrit Purusottama pourrait
done etre hauteur de cette grammaire prakrite. II \dvait au Bengale,
ce qui rend comprehensible I'existence de ce MS a la Bibliotheque
de Khatmandou ; il etait bouddhiste et le MS nepalais debute par
une invocation au Bouddha : name buddhaya. La date de Purusottama
est discutee. Dans 1' introduction au NdndrthasaUgraha il est expresse-
ment dit que Purusottama vivait au XIIP siecle.® Eajendralal
Mitra ainsi que Harapra.sad Shastri ont voulu situer Purusottama
au ou XP siecle, mais ceci semble impossible pui.sque un des
ouvrages que Purusottama cite dans sa Prayogaratnamdld est le
Vih'aprakdm ecrit en 1111 a.d.®
^ Iy\8titutiones linguae prarriticae, Bonn, 1837, p. 4.
2 “ Fur ein jungeres .Alter dagegen spncht, dass Ki, wie die spateren Grammatiker,
viel mehr Dialekte nennt als He.,” Grarnmatik der Prdkrit-Sprachen, p. 41.
^ Cf. ia preface au Trikanda^em de Purusottama edite par C. A. Seelakkhandha,
Bombay, 1016. p. 5.
* Sahitya Parisat Patrika^ vol. 39, pp. 1—6 (en Bengali).
Th. Zachariae, Citate in Kramadisvara' 8 Samlshiptasdra. BB, v, 1880, note a la
•n 9ft ■* 7)7 7
GRAMIIAIRIENS TARDIFS ET DIALECTES DF PRAKRIT
683
Quelle que soit au juste I'identite de son auteur, il reste assure
que ce traite de grammaire prakrite ecrit en 1265 connait autant
de dialectes qu'en connait Markandeya et que, comme lui, il les
repartit en quatre grands groupes : bhdm, vibhdsa, apahhramsa et
paisdcika. Voici rekabilite le temoignage de Markandeya.
Mais il y a de plus. Cette division en bhdsd et vibJidm renionte
plus loin, au Ndtyamstm. La nouvelle edition de cet ouvrage, qui
comprendra quatre volumes, mais dont les deux premiers out seuls
paru jusqu’a present, nous donne le texte au moins tel que Abliinava-
gupta I’a connu (XI®-X® siecle), sinon tel que Bharata I'a redige.^
Or les prakrits adinis dans les drames au temps d’Abhinava etaient
beaucoup plus nombreux que nc le permet Hemacandra (1088-1172)
et se repartissaient en bhdm et vibhdsd, exactement comme chez
Markandeya, Ramasarman et Purusottama. Abliinava s’efforce meme
de definir ces deux termes (vol. ii, p. 376) : bhdsd samskrtdpabhramsah \
bhdsdpabhrarnsas tu vibhdsd. “ Une bhd-sd est une corruption du
Sanskrit, une vibhdsd la corruption d'une bhdm.'’
Les “ grammairiens orientaux ” en reprenant cette ancienne
classification ont du I'elargir, non pas — comme dit Jacobi- — ‘‘ pour
I’accorder avec les nouveaux systemes de langues, " mais simplement
parce que leurs traites ne se bornaient plus a I’etude des prakrits
dramatiques et que d'autres langues etaient nees a la litterature lyrique
ou religieuse.
Car, il n'est pent etre pas inutile d'insister sur ce point, ce n’est
pas I’observation directe de langues vivantes que nous transmettent
les grammairiens prakrits de toutes les ecoles. mais les regies necessaires
et suffisantes pour ecrire des langues conventionnelles sinon artilicielles.
La date de la reaction d'un ou\’Tage grammatical n'a done pas une
grande importance et ne modifie pas necessairenient la valeur de son
temoignage. Jja seule chronologic qui puisse nous interesser est pour
ainsi dire relative et interieure Ce sont les ecoles et les sources
qu’il faut limiter et etudier. Un grammairien tardif comme Markandeya
peut done — ainsi que l avait bien pre^ui Sir George Grierson nous
conserver le reflet d'une doctrine ancienne et traditionnelle.
' Xutyasd-'^tra. with the commentary of Ahhinavagupta, ed. by M. Ramakrishna
Kavi, in four volumes : Gaekwad’s Oriental Series, Baroda.
- “ Bhavisatta Kaha,” von Dhanavala. Ahk. der Kais. Bayer, Akad. der
Bhilos.-philol. und histor. Klafi'Se, xxix Bd., 4 Abh., p. 70*.
The Expressions for “The year eonsists of twelve
months” and the like in Vedic Prose
By Hanns Oeetel
§ 1. The sentence “ The year consists of twenty-four half-
months ’ maj' be rendered in Vedic prose in three different ways : —
A. The nominatival tj-pe : caturvihsatir ardhamascih nainvut-
sarah or, with resumptive pronoun (cf. my Syntax of Cases in the
Narrative and Descriptive Prose of the Brahmanas, i, § 3, p. 2. and
§§ 8 f., pp. 10 IF.), caturvihsatir ardhamdsdh sa samvatsaralid
B. The genitival t3rpe ; catundhsatih samratsarasydr-
dhamdsah.
C. The adjectival type; caturvinsatyardhamdsah sam-
vatsarah.
Note. — A compositional expression caturvihsatih samvatsardr-
dhamdsdh does not occur though it would be theoretically possible
in view of instances like sahasram haita ddityarasmayah ^ JUB. 2. 6. 10
parallel to bahavo hy eta ddityasya rasmayah JUB. 2. 9. 10, sastis ca
ha vai trini ca satdny ddityasya rasmayah SB. 10. 5. 4. 4.®
§ 2. I. The Nominatival Type
(a) dvddasa mdsdh samvatsarah
dvadasa mdsdh samvatsarah SB. 6. 2. 1. 28 ; 36 ; 6. 2. 2. 5 : 12 ;
6. 6. 1. 5 ; 14 ; 7. 2. 2. 16 ; 7. 2. 4. 4 ; 9 ; 16 ; 22 ; 7. 3. 1. 47 ;
9. 2. 3. 6 ; 9. 3. 1. 1 ; 9. 3. 2. 8 ; 9. 3. 3. 13 ; 9. 5. 1. 38 : 10. 1. 4. 8 ;
13. 1. 2. 1 ; 13. 3. 3. 8 ; 13. 4. 1. 5 ; 13. 5. 1. 4 ; 13. 5. 3. 11 ; TS.
2. 5. 11. 6 ; 5. 2. 5. 5 ; 5. 4. 7. 6 ; 5. 4. 8. 2 ; 6 ; 5. 6. 2. 2 ; 5. 6. 7. 1 ;
3 ; 5. 6. 10. 3 ; TB. 3. 8. 1. 1 ; 3. 8. 21. 1 ; 3. 9. 18. 2 ; TA. 2. 8. 1 ;
5. 4. 10 ; 5. 6. 2 ; MS. 1. 4. 14 (64, 5-6) ; 1. 5. 6 (74, 16) ; 1. 10. 5
(145, 14) ; 2. 5. 10 (61, 5-6) ; 3. 2. 2 (16, 13) ; 3. 2. 4 (21, 5-6) ;
3. 2. 5 (21, 13) ; 3. 3. 3 (35, 13-14) ; 3. 4. 1 (41, 8-9) ; 3. 4. 3 (48, 3) ;
3. 4. 6 (52, 15) ; 3. 4. 10 (57, 15) ; 3. 8. 10 (110, 15) ; 3. 10. 2 (131,
^ This sentence is not quotable but formed by analogy to dvadasa mdsah sa
samvatsarah (see below, § 2, g),
^ The compound dditrjarasmi- is wanting in pw.
® Cf. also the compositional dve vai puru.^akapdle (read so with the MS. O), KB. 30.
4 (143, 24 ed. Lindner).
686
H. OERTEL
14) ; 4. 1. 12 (15, 13) ; 4. 3. 2 (41, 4-5) ; 4. 4. 7 (58, 19) ; 4. 4. 9 (60,
17) ; 4. 5. 7 (73, 13-14 and 74, 8) ; 4. 6. 6 (88, 12) ; 4. 6. 7 (89, 5-6) ;
K. 8. 8 (91, 13) ; 19. 11 (12, 20-21) ; 20. 3 (21, 2 and 14-15) ; 21. 5
(42, 2 and 43, 10) ; 21. 11 (51, 12) ; 21. 12 (52, 19) ; 23. 8 (84, 17) ;
26. 1 (122, 1) ; 28. 2 (154, 8) ; 29. 8 (177, 20) ; 33. 4 (29, 8-9) ; 34. 9
(43, 4) ; 35. 20 (67, 9) ; 36. 11 (77, 17) ; Kap. 7. 2 (73, 15) ; 31. 1
(147,8); 31. 5 (152, 6 and 19) ; 31. 20 (168, 24-169, 1 and 169, 11-12) ;
40. 4 (228, 2-3) ; 44. 2 (256, 22-23) ; PB. 3. 12. 3 ; 4. 2. 12 ; 4. 4. 11 ;
6. 3. 3 ; 12. 4. 17 ; 14. 1. 10 ; 16. 4. 12 ; 13 ; 16. 6. 10 ; 18. 2. 4 ;
18. 9. 7 ; 21 : 19. 5. 5 ; JB. 1. 27 i ; 135 (Caland, § 31, p. 37, 5, from
bottom) : 137 ; 179 (Caland, § 63, p. 70, 3) ; 206 ; 2. 83 (Caland,
§ 130, p. 146, 6) ; 91 ; 92 ; 93 ; 97 ; 107 ; 162 (bis) ; 163 ; 177 ;
183 (Caland, § 144, p. 178, 4) ; 199 ; 225 (bis) ; 239 ; 364 ; 376 ;
380 ; 383 ; GB. 1. 4. 19 (108, 3 ed. Gaastra) ; Vadhula S. (Caland,
Acta Oricntalia, vi, 183, 1-2 ; 196, 23 ; 216, 16 ; 233, 11 ; 234, 23).
dvddasa cai >msdh samvatsarah AB. 1. 13. 33 ; 1. 28. 41 ; 1. 29. 24 ;
2. 39. 7 ; 6. 19. 7 ; KB. 1. 1 (2, 3) ; 7. 10 (34, 9) ; 8. 1 (34, 16) ; 9. 2
(41, 21) ; 9. 4 (42, 16) : 13. 9 (61, 1) ; 14. 1 (62, 3) ; 14. 2 (62, 17) ;
16. 11 (74, 5) ; 25. 15 (119, 21) ; JB. 2. 362 (bis). dvddasa ha vai
mdsdh samvatsarah GB. 2. 1. 1 (144, 2) ; 2. 6. 1 (244, 1).
(6) trayodasa mdsdh samvatsarah
trayodasa mdsdh sami'atsarah !§B. 6. 6. 3. 16 ; 6. 7. 1. 28 ; 7. 1. 1.
32 : 7. 2. 3. 9 ; 8. 6. 3. 12 ; 9. 1. 1. 16 ; 9. 3. 3. 9 ; 13. 5. 1. 15 ; 13.
8. 3. 7 : TS. 5. 6. 7. 1 ; MS. 1. 11. 8 (170, 1) ; 3. 4. 2 (46, 19) ; K.
14. 8 (207, 11-12) ; 21. 5 (42, 11-43. 1) ; 34. 9 (43, 5) ; Kap. 31. 20
(169, 2) : I’adhfila S. (Caland, Acta Orient., vi, 183, 3 ; 234, 24).
(c) paiica (etc.) rtavah samvatsarah
jHincu rtavah samvatsarah SB. 6. 3. 1. 25 ; 6. 5. 1. 12 : 6. 8. 1. 15 ;
7. 1. 1. 32 ; 7. 2. 3. 4 ; 9 : 7. 4. 1. 34 ; 8. 6. 3. 12 ; 8. 7. 4. 9 ; 9. 2.
1. 10 ; 9. 2. 3. 41 ; 9. 4. 2. 24 ; 26 ; 9. 4. 4. 14 ; JB. 2. 291 ; Vadbula
5. (Caland, Acta Orient., vi, 233. 11-12). panca rd rtavah sam-
vat'^arah TB. 2. 7. 10. 2.
■sad rtavah samvatsarah SB. 6. 3. 2. 10 ; 6. 4. 2. 10 ; 6. 5. 4. 9 ;
6. 7. 1. 24 ; 27 ; 7. 3. 1. 35 ; 12. 2. 2. 18 ; 13. 1. 5. 6 ; 13. 5. 4. 28 ;
13. 8. 2. 6 ; TA. 2. 8. 1 ; PB. 19. 18. 5 ; JB. 2. 420 (Caland. § 168,
p. 219, 1 from bottom). .sad vd rtavah sariivatsarah TS. 5. 2. 6. 1 ;
Passages from JB. for which no references in parentheses are given are not vet
published.
EXPRESSIONS FOR “ THE YEAR CONSISTS OF TWELVE MONTHS " 687
5. 4. 2. 2 ; 5. 6. 7. 1 ; MS. 1. 7. 3 (111, 19) ; 3. 4. 6 (52, 12M3) ; K.
9. 1 (104, 11) ; 20. 4 (21, 20) ; 21. 5 (42, 18 ; read so with the MSS.
St. and \V la and the Kap. parallel instead of von Schroeder's sad
vd rtava rtavas samvatsarah, cf. K. 22. 6 [62. 18] where the MS. Ch.
wrongly reads rtavas twice) ; 22. 6 (62, 18) ; Kap. 8. 4 (83, 7) ; 31. 6
(152, 24) ; 31. 20 (168, 22) ; 34 (176, 17) ; KB. 15. 2 (66, 1) ; 19. 7
(87, 1) ; 23. 1 (102, 21) ; 26. 16 (128. 20) ; ,30. 10 (146. 11) ; JB. 2.
142 ; 199 ; 431 (Caland, § 169, p. 229, 1-2) ; 3. 61 ; Vadhida S.
(Caland, Acta Orient., vi, 196, 24).
sapta rtavah sariivatsarah SB. 6. 6. 1. 14 ; 6. 6. 2. 7 ; 6. 8. 2. 7 ;
7. 3. 2. 9 ; 9. 1. 1. 26 ; 9. 1. 2. 31 ; 9. 2. 3. 45.
(d) caturvihsatir ardhatnasdh sariivatsarah
caturvihsatir ardhamdsdh samvatsarah TS. 5. 1. 8. 5 ; 5. 6. 7. 2 ;
TA. 2. 8. 1 ; PB. 4. 4. 1 ; 6. 3. 4 ; 14. 1. 11 ; 15. 1. 10 ; SB. 3. 1. 21 ;
\adhula S. (Caland, Acta Orient., vi, 233, 15-16). caturvihsatir
vd ardhamdsdh samvatsarah MS. 3. 4. 6 (-52, 4, and 17).
(e) dvau rndsdv Huh
dvau hi rndsdv Huh §B. 7. 4. 2. 29 ; 8. 2. 1.16; 8, 3. 2. 5 ; 6 ;
8. 4. 2. 14 ; 8. 7. 1. 6. dvau-dvan hi rndsdv Huh PB. 10. 2. 8.
(/) trihsad rdtrayo mdsah
trihsdt-trihsad vdi rdtrayo mdso, yo mdsah sd samvatsarah MS.
1. 10. 8 (148, 8) ; K. 36. 2 (70. 6-7 where . . . yo vai mdsah . . .) ;
contrast K. 34. 9 (43, 13) trihsan mdso rdtrayah, below § 3, II
(Genitival Type), /.
AVith resumptive demonstrative pronoun : —
{g) dvddasa mdsah sa samvatsarah
dvddasa mdsdh parica rtavah sa sarnvatsarah TS. 5. 6. 7. 2 ; 6. 3.
7. 1. panca rtavo dvddasa mdsd e.sa sarnvatsarah MS. 1. 7. 3 (111,
14) ; 3. 7. 2 (79, 7-8) ; 4. 3. 2 (41. 6). dvddasa mdsah parica rtavah
sa vai sarnvatsarah PB. 18. 2. 14 ; 18. 4. 11 ; 18. 9. 5. parica rtavo
dvddasa mdsd esa vdva sa sar'nvat.sarah K. 9. 1 (104, 6) ; Kap. 8. 4
(83, 2). Similarly sastis ca ha vai trini ca saidny etac chatarudriyarn
SB. 9. 1. 1. 43 (s. below § 6 in fine).
AVith resumptive tdvdn : —
{h) dvddasa mdsds tdvdn sarnvatsarah
dvddasa rndsdh parica rtavas tdvdn sarhvatsarah AB. 1. 16. 44.
688
H. OERTEL
§ 3. II. The Gbnitival Type
(a) dvadasa masdh samvatsarasya
dvddasa vai mdsah samvatsarasya SB. 1. 2. 5. 13 (= Kanva 2. 2. 3.
11) ; 1. 3. 5. 10 ; 11 ; 2. 2. 2. 4 (= Kanva 1. 2. 2. 3) ; 3. 4. 4. 19 ;
3. 6. 4. 23 ; 4. 3. 1. 5 ; 4. 6. 1. 11 ; 5. 2. 1. 2 ; 5. 2. 5. 15 ; 5. 3. 5. 5 ;
5. 4. 5. 14 ; 20 ; 5. 5. 2. 2 ; 5. 5. 5. 7 ; 11. 2. 6. 11 ; 11. 5. 4. 9. ; 12.
2. 1. 8 ; 12. 2. 2. 6 ; 12. 7. 2. 19 ; 14. 2. 2. 12. (ZmtZasa
samvatsarasya SB. 11. 6. 3. 8 ; 12. 2. 3. 6 ; 12. 3. 2. 2 ; GB. 1. 5. 5
(119, 5). tasya (scil. samvatsarasya) dvddasa mdsah paiica rtavah
SB. 8. 4. 1. 11-13. tasya (scil. saynvatsarasya) dvddasa vidsdh sad
rtavah SB. 8. 4. 1. 14. tasya (scil. samvatsarasya) dvddasa mdsah
sapta rtavah SB. 8. 4. 1. 15; 16.
(6) trayodasa mdsdh samvatsarasya
trayodasa vai mdsdh samvatsarasya ^B. 3. 6. 4. 24 ; 14. 1. 3. 27 ;
14. 3. 2. 16 ; GB. 1. 5. 5. (119, 6). tasya (scil. samvatsarasya)
trayodasa mdsdh sapta rtavah SB. 8. 4. 1. 17.
A combination of II (a) and (b) in : —
dvddasa vd vai trayodasa vd samvatsarasya mdsdh ^B. 2. 2. 3. 27
(= Kanva, 1. 2. 3. 23) ; 5. 4. 5. 23 ; 5. 5. 5. 19.
(c) panca (etc) rtavah samvatsarasya
traya rtavah samvatsarasya SB. 12. 3. 2. 1. trayo vd rtavah
samvatsarasya SB. 3. 4. 4. 17 ; 11. 5. 4. 10 ; GB. 1. 5. 5 (119, 1).
panca rtavah samvatsarasya SB. 1. 5. 2. 16 ; 1. 7. 2. 8 ; 2. 1. 1. 12
(= Kanva 1. 1. 1. 8) ; 3. 1. 4. 20 ; 3. 6. 4. 18 ; 11. 7. 4. 4 ; 12. 2. 2.
19 ; 12. 3. 2. 1 ; 14. 1. 2. 14. panca vd rtavah samvatsarasya SB.
3. 1. 3. 17 ; 3. 1. 4. 5 ; 3. 3. 3. 5 ; 3. 4. 1. 14 ; 3. 9. 4. 11 ; 4. 1. 1.
16 ; 4. 5. 5. 12 ; 5. 1. 2. 9 ; 14. 1. 1. 28. tasya (scil. saihvatsarasya)
dvddasa mdsdh paiica rtavah SB. 8. 4. 1. 11-13.
sacl rtavah sainvatsarusya SB. 3. 6. 4. 19 ; 12. 3. 2. 1. sad vd
rtavah samvatsarasya SB. 1. 2. 5. 12 (= Kanva 2. 2. 3. 11) ; 2. 2. 2. 3
(= Kanva 1. 2. 2. 2) ; 3. 4. 4. 18 ; 4. 2. 2. 7 ; 4. 4. 5. 18 ; 4. 5. 5. 12 ;
5. 2. 1. 4 ; 11. 5. 4. 7 ; 10 ; GB. 1. 5. 5 (119. 2-3). sad eva rtavah
samvatsarasya SB. 2. 1. 1. 13 (bis ; the Kanva parallel 1. 1. 1. 9 sad
vd rtavah samvatsarasya and sad rtavah samvatsarasya) ; 4. 5. 5. 12.
tasya (scil. samvatsarasya) dvddasa mdsdh sad rtavah SB. 8. 4. 1. 14.
sapta rtavah samvatsarasya SB. 12. 3. 2. 1. sapta vd rtavah
samvatsarasya GB. 1. 5. 5 (119, 4). tasya (scil. samvatsarasya)
trayodasa mdsdh sapta rtavah SB. 8. 4. 1. 17. tasya (scil. snin-
vatsarasyn) dvddasa mdsdh sapta rtavah SB. 8. 4. 1. 15 ; 16.
EXPRESSIONS FOE “ THE YEAR CONSISTS OF TWELVE MONTHS ” 689
(d) caturvinsatih samvatsarasyanUiamasah
caturvihsatih samvatsarmyardhamdsah K. 21. 5 (43, 4) ; 22. 1
(57, 16) ; 33. 2 (28, 3-4) ; 33. 3 (29, 8) ; Kap. 31. 20 (169, 5) ; PB.
16. 7. 5 ; 23. 21. 3. caturvinmtir vai sarhvatsarasydrdhamdsdk
SB. 2. 2. 2. 5 (the Kiinva parallel 1. 2. 2. 4 caturvihsatir vd ardhamd-
sdh samvatsarasya) ; 4. 1. 1. 15 ; 4. 6. 1. 12 ; 5. 4. 5. 21 ; 11. 5. 4. 8 ;
KB. 9. 6 (44, 7) ; 19. 8 (87, 7). caturvinsatir ardhanidsdh sam-
vatsarasya GB. 1. 5. 5 (119, 7-8). caturvihsatir vd ardhamdsdh
samvatsarasya SB. Kanva 1. 2. 2. 4 (the Madhyandina parallel 2. 2. 2.
5 caturvihsatir vai samvatsarasydrdhamdsdh). tasya (sell, sam-
vatsarasya) caturvihsatir ardhamdsdh SB. 8. 4. 1. 18 ; 19 : 21-24.
sadvihsatir ardhamdsdh samvatsarasya GB. 1. 5. 5 (119, 9).
tasya (sell, samvatsarasya) sadvihsatir ardhamdsdh SB. 8. 4. 1. 25.
(e) vihsatisatam rtor ahdni
vihsatisatam vd rtor ahdni K. 11. 7 (51, 23) ; 30, 10 (146, 4).
(/) trihsan mdsasya rdtrayah
trihsan mdsasya rdtrayah SB. 9. 1. 1. 43 ; 10. 4. 2. 23 ; 24 ; AB.
3. 41. 2 ; JB. 2. 109 ; 375 ; and so with mdsah genitive sg. to mds-
K. 34. 9 (43, 13) trihsan mdso rdtrayah. For trihsat-trihsad vai rdtrayo
mdsah MS. 1. 10. 8 (148, 7) ; K. 36. 2 (70, 6-7) see above,
§2,1 (Nominatival Type), /.
(g) dve sarhvatsarasydhardtre
dve vai samvatsarasydkordtre SB. 12. 3. 2. 1. dve ahordtre
samvatsarasya GB. 1. 5. 5 (118, 14).
{h) paFicadasdrdhamdsasya rdtrayah
paheadasdrdhamdsasya rdtrayaJi MS. 1. 7. 3 (111, 17) ; K. 9. 1
(104, 9) ; 21. 5 (43, 2) ; 33. 8 (34, 22) ; 34. 9 (43, 6-7) ; Kap. 8. 4
(83, 4-5) ; 31. 20 (169, 3) ; PB. 4. 2. 8. paheadasa vd ardhamdsasya
rdtrayah SB. 1. 3. 5. 8 ; TS. 2. 5. 8. 3 ; 5. 6. 7. 2 ; TB. 3. 3. 7. 1 ;
3. 9. 11. 2 ; JB. 1. 132 ; 2. 109 ; 375 ; Yadhula S. (Caland, Acta
Orient., \d, 233, 10). pahcadasdpuryamdnasydrdhamdsasya rdtrayah
JB. 1. 251. pahcadasdpochato ^ (so and °porchato. °poccato the MSS.)
rdhamdsasya rdtrayah JB. 1. 251.
^ The use of the present participle of the root 1 upu to designate the
waning moon (= apak.nyamnna- ; 8B. 10. 4. 2. 17 pancadasapuryamanasya rupani
pahcadampah^iyamdriasya) i.s noteworthy and does not appear to occur elsewhere.
At Vadhula S. (Caland. .4r/a Orient, vi, 133, 4. and 134, 11). arachannnpnrapaksa-
is used in this sense in contrast to abhipurya-mdnn\^pakm-'\.
690
H. OEKTEL
(*) pancadasa piirvapaksaparapaksayor ahani
pancadaki vai purvapaksdparapayor ahdni KB. 3. 2. (9, 7-8).
(j) astdvinsatis ca satduy asTtis ca samvatsarasya pdddhds ca
pddardtrayas ^ ca
astdvinsatis ca ha vai satdny asitis ca samvatsarasya pdddhds ca
pddardtrayas ca GB. 1. 5. 5 (119, 17-120, 1).
{k) caturdasa ca satdni catvdrihsac ca sanivatsarasydrdhdhds
cdrdhardtrayas ^ ca
caturdasa ca ha vai satdni catvdrinsac ca sanivatsarasydrdhdhds
cdrdhardtrayas ca GB. 1. 5. 5 (119, 14-15).
{1) pancadasdnam varsdndm trlni ca satdni sastis ca paurnamdsyds
cdntdvdsyds ca
tesdm pancadasdnam varsdndm trini ca satdni (11 trini caiva satdni)
sastis ca paurnamdsyds cdmdvdsyds ca SB. 11. 1. 2. 10 ; 11.
[m) dasa ca sahasrdny astau ca satdni samvatsarasya muhurtdh
dasa ca vai (GB. ca ha vai) sahasrdny astau ca satdni samvatsa-
rasya muhurtdh SB. 12. 3. 2. 5 ; GB. 1. 5. 5 (120, 2-3).
(n) sapta ca satdni vihsatis ca sariivatsarasydhordtrdni
sapta ca vai satdni vihsatis ca samvatsarasydhordtrdni ^B. 12. 3. 2.
4. tasya vd etasya samvatsarasya prajdpateh sapta ca satdni vihsatis
cdhordtrdni jyotlhsi SB. 10. 4. 2. 2 {prajdpateh is apposition to sam-
vatsarasya and jyotlhsi to ahordtrdni). sapta ca vai vihsatisatdni
ca samvatsarasydhordtrdndm KB. 11. 7 (52, 2). sapta ca ha vai
satdni vihsatis ca samvatsarasydhdni ca rdtrayas ca GB. 1. 5. 5 (119,
12-13). sapta ca vai satdni vihsatis ca saihvatsarasydihordtrdh
AA. 3. 2. 1 (]33, 6 ed. Keith). saptavihsatisatdni samvatsarasyd-
hordtrdndm SA. 8. 1 (312, 13-14 in Keith's ed. of AA.).
[o) trini ca satdni sakis ca samvatsarasya rdtrayah
trini ca vai satdni sastis ca samvatsarasya rdtrayah SB. 11. 1. 2.
10 ; 12. 3. 2. 3. trini ca vai satdni sastis ca say'nvatsarasydhdm
SB. 1. 3. 5. 9 ; 11. 1. 2. II ; 12. 3. 2. 3. trini ca ha vai satdni sastis
ca samvatsarasydhordtrdni GB. 1. 5. 5 (119, 10-11). sa-^is ca ha
vai trini ca satdni samvatsarasya rdtrayah SB. 10. 4. 3. 13 ; 10. 5.
4. 10. sastis ca ha vai trini ca satdni samvatsarasydhdni SB. 10. 4.
3. 19 ; 10. 5. 4. 10. saMis ca vai trini ca satdni samvatsarasydhdni
^ The compounds slxhX padaratri- are wanting in pw
^ The compounds ardhdha- and ardkanUri- are wanting in pw.
EXPRESSIOXS FOR THE YEAR CONSISTS OF TWELVE MONTHS 691
K. 33. 2 (28, 5-6). trlni vai sastisatani sariivatsarasyahnam KB.
3. 2 (9, 11-12) ; 11. 7 (52, 1) ; 16. 9 (73, 13) ; 19. 8 (87, 8).
(p) tesdm (scil. samvatsardndm) sattrihkit purnamdsdh
ye vai trayah samvatsards iesdtii sattrihsat purnamdsd, yaii dvau
tayos caturvihsatih MS. 1. 10. 8 (148, 17-18) ; K. 36. 3 (70. 20-71, 1).
{q) aparimitdh samvatsarasya rdtrayah
aparimitds (MS. aparimitd vai) samvatsarasya rdtrayah MS. 1. 10. 7
(157, 3-4) ; K. 36. 11 (78, 10).
(r) ydva)tti (tdranti) samvatsarasydhdni
tdvanti samvatsarasydhdni SB. 9. 1. 1. 43 ; 9. 3. 3. 18 ; MS. 1. 7. 3
(111, 17-18) ; K. 9. 1 (104, 9-10) ; Kap. 8. 4 (83, 5-6) ; AB. 4. 12. 7 ;
9- td ydvatih samvatsarasya rdtrayah K. 36. 2 (70. 9-10).
tdvatih samvatsarasya rdtrayah TS. 2. 5. 8. 3; MS. 1. 10. 8
(148, 10-11). tdvatyah samvatsarasya rdtrayah PB. 4. 2. 7 ; 9. 3.
tdvatis samvatsarasya rdtrayah JB. 2. 163 ; 212 ; 375,
tdvatJr rto rdtrayah JB. 2. 212. etdvanti hi samvatsarasydhordtrdni
SB. 7. 3. 1. 43. etdvanti vai samvatsarasydhordtrdni MS. 4. 5. 3
(67, 12-13). tdvanti samvatsarasydhordtrdni JB. 2. 238 (Caland,
§ 147, b ; p. 185, 14 from bottom) ; 240. ydvanti vai samvatsarasyd-
hordtrdni SB. 6. 2. 2. 29. ydvanti samvatsarasydhordtrdni K. 15.
10 (217, 2) ; PB. 18. 11. 6. tdvantas samvatsarasydhordtrdh JB.
2. 204. etdvanto hi samvatsarasya mnhurtdh SB. 10. 4. 3. 20.
tdvantah sahasrasarhvatsarasya mnhurtdh ^B. 10. 4. 4. 2. — — etdvanto
vai pahcdndm samvatsardndm ahordtrd ydvatlr eta istald ydvdn agnih
samcitah Vadhula S. (Caland, Acta Orient., vi, 239, 25).
(s) hati samvatsarasydhdni
kati samvatsarasydtirdt rdh SB. 12. 2. 1. 6. kati te pitdi sam-
vatsarasydhdny amanyata SB. 12. 2. 2. 13 ; GB. 1. 4. 24 (111, 14-
112, 1) ; JB. 2. 431 (Caland, § 169, p. 228, 7 from bottom). kati
samvatsarasydhdni pardnci SB. 12. 2. 3. 13 ; GB. 1. 4. 16 (106, 10-11).
§ 4. III. The Adjectival Type
(a) sadrtuh ^ samvatsarah KB. 14. 1 (61. 12) ; 20. 3 (91, 1).
(b) caturvinsatyardhamdsah ^ samvatsarah SB. 6. 2. 1. 21 ; 10.
4. 2. 18 ; JB. 1. 212 ; 2. 4 ; 91 (bis) ; 92 (bis) ; 93 (ter) ; 97 ; 107 ;
* The adjectives sadrtu- and caturvinhatyardhamasa- are wanting in pw. Cf.
mtamrada- in the Mantra TS. 5. 7. 2. 4 (etc.) tesdm rtundm mtasdraddndm.
692
H. OERTEL
111 ; 119 ; 124 (Caland, § 137, p. 162, 12) ; 127 ; 162 (bis) ; 163 ;
176 ; 177 ; 225 (bis) ; 239 (bis) ; 282 ; 285 ; 308 ; 349 ; 359 ; 375 ;
3. 173 ; JUB. 3. 38. 9. mturviiisatyardhamdso vai samvatsarah
JB. 2. 377 ; 380 ; 410 (Caland, § 167, p. 218, 10 from bottom) ;
412; 435.
§ 5. St'jrMARY
(а) Outside of JB. the adjectival type ^ (above § 4) is very
rare : sadrtu- occurs twice in KB., caturvinsatyardhamasa- twice iu
SB. (once each in books vi and x). But in JB. the adjectival
caturviiisatyardhamdsas samvatsarah is the only expression for “ The
year consists of twenty-four half-months ’’ (thirty-six times in JB.,
and once in JUB.).
(б) The nominatival dvddasa mdsah samvatsarah and trayodasa
rndsdh samvatsarah (above, § 2, a and b) is in SB. confined to books
vi-x and xiii (eight times each in books vi and vii, once in book Uii,
seven times in book ix, once in book x, and five times in book xiii).
In the remaining books SB. uses the genitival (above, § 3, a and
b) dvddasa rndsdh samvatsarasya, trayodasa mdsah saihvatsarasya,
dvddasa vd trayodasa id samvatsarasya mdsah (three times in book i,
twice in book ii, three times in book iii, twice in book iv, nine times in
book V, seven times in book viii, three times in book ix, five times in
book xii, and three times in book xiv).^ It is noteworthy that this
genitival expression is not found in any of the other Brahmanas ;
all of them, with the single exception of GB. 1. 5. 5 (119, 5 and 6),
which depends on ^B. 12. 3. 2. 1 £f., use the nominatival expression
(above, § 2, o and b) ; cf. below, § 5, c in fine.
(c) The nominatival paiica (etc.) rtavah samvatsarah (above,
§ 2, c) occurs in SB. in books \'i-ix and xiii (eleven times in book vi,
six times in book vii, twice in book viii, eight times in book ix, and
three times in book xiii), and strangely enough also once in book xii,
which elsewhere (four times) has the genitival expression. In the
remaining books SB. uses the genitival (above, § 3, c) panca
(etc.) rtavah samvatsarasya (three times in book i, four times in book ii.
* Cf. drikapalam hi sirah, SB. 8. 4. 4. 4.
“ Cf. for a distantly similar contrast of nominative and genitive, JB. 2. 77 {JAOS.
XV, 240, 6), katama aditya iti, dvddasa masds samvatsara (so all the MSS.) iti horhcnita
ddityd, ete hldarh sarvam adadand yanti, tasmad ddityd iti, but the parallels SB. 11.
6. 3. 8 and 14. 6. 9. 6, dvddasa masdh samvatsarasya : “ ‘ Who are the Adityas ? ’
The twelve months, the year (so JB. ; the twelve months of the year SB.). these are
the Adityas.’ ”
EXPEESSIOXS FOE “ THE YEAE CONSISTS OF TWELVE MONTHS ” 693
nine times in book iii. six times in book iv, twice in book v, four times
in book viii, three times in book xi, four times in book xii [but note
once in this book the nominatival expression], and twice in book xiv).
This genitival expression is foreign to all the other Brahmanas (cf.
above, § 5, 6 in fine) ; all of them, with the single exception of GB.
1. 5. 5 (119, 2-3 and 4) which depends on SB. 12. 3. 2. 1 ff., use the
nominatival expression (above, § 2, c).
(d) For “ The year consists of twenty-four half-months ” SB.
has only the genitival expression caturvihsatiJi samvatsarasydr-
dhamdsah (once in book ii, twice in book iv, once in book v, seven
times in book viii, and once in book xi). The absence of the
nominatival expression in SB. is due to the fact that books vi-x
and xiii for which the nominatival construction is characteristic
(above, § 5, 6 and c) have no occasion to express the thought.
Of the other Brahmanas K. (four times), Kap. (once), KB. (twice)
and GB. (twice) show the genitival expression only, while TS. (twice),
TA (once), MS. (twice), and SB. (once) know only the nominatival
construction. PB. uses the nominatival expression once each in books
iv, vi, xiv, and xv, but the genitival construction once each in books
xvi and xxiii.
(e) For ‘‘A season consists of two months” SB. has the
nominatival draw mdsdv rtiik six times (once in book vii, five times
in book viii, all books for which the nominatival construction is
characteristic, see above, § 5, 6 and c). PB. has it once (in book x).
But K. has the genitival vihsatisatam rtor ahdni (twice) “A
season consists of 120 days’’.
(/) In general the genitival type (above, § 3, as) is far more
frequent than the nominatival type (above, § 2, a^) ; the latter
is confined to the expressions dvadasa niasdh sariivatsarak, trayodasa
tndsdh samvatsarah, panca (etc.) rtavah samvatsarah, caturi'insahr
ardhamasdh sarntatsarah, dvaii mdsdv rtxh, and (above, § 2, /)
trihsad rdtrayo mdsah.
§ 6. The same variation between nominative and genitive is
found in the expressions for “ The Agnistoma consists of twelve
Stotras ” and “ The night(-rite) consists of twelve Stotras . M e have,
on the one hand, the genitival dvddasdgnistonmsya stotrdni K.
26. 1 (122, 1) ; Kap. 40. 4 (228. 2) ; TB. 1. 2. 2. 1 ; JB. 1. 179 (Caland,
§ 63, p. 70, 3) ; 206 ; and dvadasa vd aynistomasya stotrdni dvadasa
694 EXPEESSIOXS FOR “ THE YEAR COXSISTS OF TWELVE MOXTHS''
rdtreh JB. 1. 206 ^ ; on the other hand, the nominatival
dvddasa stotrany agnistomah PB. 4. 2. 12 ; 6. 3. 3 ; and dvddasa
stotrdny agnistorno dvddasa stotrdni rdtrih PB. 9. 1. 24.^ Cf. with
resumptive pronoun, sastis ca ha vai trini ca satdiiy etac chatarudriyam
SB. 9. 1. 1. 43 “ The Satarudriya(-litany) consists of 360 (formulas) ”.
Note. — ^For the sake of completeness I add the following genitival
examples : jtancadasdndm u vai gdyatflndm trini ca satdni sastis
cdl'sardni SB. 1. 3. 5. 9 ; tesdm (scil. chandasdm) tisras cdsitayo
’ksardni 'paiicacatvarinsac ca SB. 10. 1. 2. 9 ; tricdni tesdm (scil.
chandasdm) sapta ca satdni vihsatis cdksardni SB. 10. 5. 4. 7 ; tdsdrh
vd etdsdrh pancdndth vydhrtlndm saptadasdksardni SB. 12. 3. 3. 3 ;
tdsdrh (scil. vydhrtlndm) saptadasdksardni SB. 1. 5. 2. 17 ; ydvanti
hi saptdndm chandasdm aksardni tdvanty etasya sadrcasydksardni
SB. 7. 3. 1. 41 ; tdsdrh (scil. ream) nava paddni SB. 6. 4. 2. 5 ; tasya
(scil. sdmnah) trirvacana (“ when it is sung thrice ”) ekavihsatih paddni
SB. 3. 1. 20; caturdasa vd etdsdrh (scil. grlvdridrn) karukardrii SB.
12. 2. 4. 10 ; dvdtrihsad vd etasya (scil. andkasya) karukardrii trayo-
dasdgries citipurlsdrii SB. 7. 1. 1. 32 ; 7. 2. 3. 9 ; 8. 6. 3. 12 ; 9. 3. 3.
9 ; sastis ca Irlrxi ca satdny anyatarasyestakd abhavann evarn anyata-
rasya SB. 10. 4. 2. 4 ; tisras-tisro 'sltaya ekaikasyestakd abhavan
SB. 10. 4. 2. 5 ; catuscatvdrihsarh satarn ekaikasyestakd abhavan
SB. 10. 4. 2. 7.3
^ Cf. for the genitival expression tn^yn (soil, atiratrasya) saMis ca triTiii ca satani
Motnyak K. 33. 2 (28, 5) ; temrh va etestarh caturxtarn ukthydnarii sahasrnrh stotriydh.
KB. 21. .“) (9n, 21) ; paheadasa hy asya (scil. ukthyasya) stotrdriii bhavanti pancadasa
sastrdni KB. 24. 9 (111, 12-13); tayor ctayoh paiicadasasaptadasayor dvdtrinsnt
stotriydh SB. 10. 1. 2. 8; tn.^ya (soil, stomnsya) nnvntimtam stotrlynh PB. 16. I. 8
tasya (scil. cknha^iya) dvisafdh stotriydh PB. 16. 9. 3 ; tasyai (scil. iste.h) paheadasa
samidhenyo hkavanti §B. 11. 4. 3. 18; 13. 4. 1. 13; 13. 4. 2. 7 ; tasyai (scil. i-steh)
saptadasa samidhenyo bhamnti 13. 4. 1. 1.5; 13. 4. 2. 9 ; 13; nnvatisatam u ha
I'd agnistomasya stotriydh Vadhula S. (Calami, Acta Orient. ^ vi, 234, 27).
2 The MS. has the locative ; dvddamgnistome stotrdni MS. 4. 5. 4 (68, 7-8, where
the MSS. H and Bb read 'gnistome^ the other MSS. Agnistorno) 4. 5. 7 (74, 9, where
all MSS. read gnistnme).
^ Examples like pisti.s ca ha vai trini ca satdni purnsnsydsthini and sastis ca ha
trini ca satdni pnrusasya majjanah 10. .5. 4. 2; trini ca satdni saMis ca purusas-
ydsthini and trini ca satdni .^nstis ca purusasya majjanah SB. 12. 3. 2. 3 ; sapta ca
satdni rinsatis ca purusasydsthini ca majjdnas ca 5B. 12. 3. 2. 4 are logically different ;
for the human bod^' consists (i.e. is wholly made up) not only of bones
and marrow but of five elements (e.g. AB. 2. 14. 7 = 6. 29. 4 pahkto 'yam puru.^ah
paheadhd vihito : lomdni trah mdnsam astki majjd) ; it can therefore properly be said
to contain bones and marrow, but not to consist of them.
Syntax of the Past Tense in Old Rajasthani
By B. S. Pandit
T)ROFESSOR J. BLOCH in L'lndo-aryeti dn Veda aux temps
modernes, p. 271 f., has discussed the passage of the passive
construction of the past tense of transitive verbs (based on the past
participle of Sanskrit) into an active construction in which the verb
agrees in number and gender with the logical subject. It is obvious
that during the course of this evolution there were periods of
fluctuation. An astonishing fluctuation is found in the Rajasthani MS.
Nala Ici bdta (R.A.S. Todd Coll. No. 81), written at some date prior
to A.D. 1806.
Construction 1
The past participle (past tense) agrees with the logical object which
is in the direct case ; the logical subject is in the instrumental in
the singular {-ai), general oblique in the plural (-«). Sometimes the
gen. obi. is used also in the singular for the instrumental ; and when this
is the same in form as the direct, the process of confusion begins.
There will thus be found confusions between Constructions I and VII.
Subject in
Object in
Verb agrees with
Instrumental.
Direct.
object.
bugalai
doho
kahijo
3a 2
tai
kai'ntha
ladhyau
21b 4
mile
bikho
liyo
16« 5
i
mdga
bycihl
446 11
tell 1
utama
junl
296 5
Darnatl ^
dohd
kahyd
436 8
Oblique plural.
guvdld
doho
kahyo
4a 7
sagald
bicdra
knryo
66 10
Construction II
The logical subject is in the instrumental, the logical object in the
direct ; the verb agrees in number and gender with the logical subject.
This seems to have been the result of the construction with intransitive
^ Direct, gen. oblic|ue and instrumental are identical in form.
696
B. S. PANDIT —
verbs together with the replacement of the use of the direct case by
instrumental (originally in pronouns) and later by the coalescence in
form of the direct and oblique. Thus rani can be in origin either direct
or instrumental. There are, therefore, confusions of Constructions II
and VIII.
Subject in Object in Verb agrees with
Instrumental. Direct. Subject.
rdyii usdsa nakhi 196 4
Damati nisdsa nakhi 37a 8
Constrtwtion III
The subject is in the instrumental, the object in the oblique,
usually with the postposition ne or na ; the verb agrees in number and
gender with the object. As has already been remarked, there is some
difficulty in distinguishing whether the subject is in the instrumental
or the direct. Thus there is confusion between Constructions III and IT .
This construction has its origin in the loss of distinction between
instrumental and general oblique. The general oblique without
postposition might be ambiguous since it might be either the logical
object or the logical subject.
Subject in Object in Verb agrees with
Instrumental. Oblique + ne.
Object.
mke
tu ne
chodiyo
236 6
sahalyd
kasidd na
sardyo
26a 8
rdjd ^
nidnasd ne
hhejyd
4a 9
rdjd ^
rdni ne
vthdi
196 7
Construction
IV
The subject is
in the direct, the object in the oblique
with
! verb agrees in number and gender
with the object.
Subject in Object in
Verb agrees with
Direct.
Oblique + ne.
Object.
.so
caravdddrd na
mdryd
41a 4
so
yhord tin
laid kiyd
41a 5
Construction V
The subject is in the direct, the object in the oblique with 7ie ;
the verb agrees in number and gender with the subject. The old
passive construction has here passed into the active.
* Direct, gen. oblique, and instrumental are identical in form.
•(
SYNTAX OF THE PAST TENSE IN OLD RAJASTHANI 697
Subject in
Object in
Verb agrees with
Direct.
Oblique + ne.
Subject.
raja
u na
buldyo
41a 6
Nala
Dholdji na
khildva chd
476 8
raja
rani ne
kahyo
26a 10
Damaiti
NaJa ne
puchl
146 12
Construction VI
Sanskrit type
in which the
subject is in the
instrumental
and the verb is the impersonal neuter singular survives in this. The
subject is either in the instrumental proper or the general oblique ;
the verb ends in -7 (< Skt. -itam). In an isolated case (tadaJco huvl
35a 3) this impersonal construction has been extended to an
intransitive verb, the subject being in the direct and the verb
(apparently not agreeing in gender) being derived from an analogical
formation from the Skt. neuter impersonal past participle in -itam.
Subject in Verb is
Instrumental. Old Neuter Singular.
mohe
jdni
45a 7
the
bicdri
486 8
umaravd
kahJ
416 11
sddhd
kahJ
56 6
Nala 1
kaki
486 8
Nala ^
jdni
66 13
Construction YII
Both subject and object are in the direct ; the verb agrees in number
and gender with the object.
Subject in Object in Verb agrees with
Direct. Direct. Object.
badhika jdla ndkhyo 46 11
badhika chala karyo 6o 7
Construction VIII
Both subject and object are in the direct, the verb agrees in number
and gender with the subject.
Subject in Object in Verb agrees with
Direct. Direct. Subject.
raja thdri surati kaJiyo 126 3
* Direct, oblique, and instrumental are identical in form.
VOL. VIII. PARTS i AND 3.
4 .')
698
SYNTAX OF THE PAST TENSE IN OLD RAJASTHANI
Construction JX
The subject is in the instrumental ; the verb is in the masculine
singular. This follows from the loss of the neuter gender. The masculine
has replaced the neuter, which here (-7 < -itain) seemed to be feminine.
Subject in
Instrumental.
Damati
ram
\'erb is
-Masculine Singular.
kahau 276 8
kahyau 286 16
Vedico yuh “se ipsum”
By Vittore Pisani
"\/r BLOOMFIELD, cerca, nella Miscellanea in onore di Ernst
• Kuhn {Aufsdtze ziir KiiJtur- und Sprachgeschichte . . . Monaco,
1916, p. 211 sgg.), di sostenere la sua proposta {JAOS., xxvii, p. 72 sgg.),
secondo cui RV., viii, 18, 13 dovrebbe leggersi :
go nah Me cid nriMati raksastvena mdrtyah
s}fd{h sd evai ririsisldyur jdnah
in luogo del tramandato ririslsta ydr ; °Ad° rappresenterebbe una
contrazione di °.fta d°, e il Bloomfield traduce in conseguenza :
“ The mortal who with demonic practices desires to harm us :
may that person by his own doings injure his life ! ’’
Kon si puo negar ragione al Bloomfield quando difende, per
ragioni grammatical! {ririslMa. aoristo raddoppiato, e causative)
alle quali si possono aggiungere quelle testuali, la lezione tradizionale
contro I'emendazione del Dizionario di Bohtlingk-Roth approvata
da Oldenberg ; svaih sd evai ririslsta dvayiir jdnah ; raa quanto egli
dice a p. 213 non mi pare bastante per dimostrare che ragioni metriche,
e cioe il bisogno d'avere una fine di verso • — w — , fossero da tanto
da ridurre ad d I'd che ci aspetteremmo come risultato dalla con-
trazione dell’ -a finale di ririslsta con quello iniziale di uyuh. Anche
ammesso che purtva jdryah RT., O, 2, 7 valga piiri iva ajurydh, qui
SI potrebbe trattare di una specie d'elisione ; e forse il principio del-
1 elisione va applicato anche in qualche altro passo del RV., ed esso
appare di data indeuropea, accanto all’ altro, generalizzatosi in ai.,
della contrazione e della formazione di dittonghi discendenti o ascen-
dent! (a -f a = d, a = i = €. i ~\- a = ya ecc.) nel caso d'incontro
di vocali nel sandhi. Nel piu antico greco la elisione ha luogo nel
sandhi sintattico (ma anche rdAAa ecc.), laddove la contrazione si
e afifermata nell’ incontro di vocali finale e iniziale dei due membri
d’un composto (arpardyos da orpaTo-ayos ) ; lascio impregiudicata
la questione, se la elisione che anche ha luogo fra i membri d'un
comporto (dAe'^-aeSpo?) sia continuazione di un fatto ie. o imitazione
di quanto avveniva nel sandhi sintattico. In latino abbiamo la si-
nalefi in cui, si pronunziasse o non la finale della prima parola, il
valore prosodico dell’ unica sillaba risultante e quello deUa seconda
vocale ; accanto alia sinalefi abbiamo forse ancora in Plauto la
700
VEDICO YVB “ SE IPSUM ”
contrazione, la dove si usa parlare di iato prosodico : ila me di ament
va letto ecc. secondo Eau (Zeitschrift fur verghicherde
Sprachforschung, Iviii, 145 sgg.), Pero, tanto in greco quanto in
latino, la prosodia della sillaba risultante da elisione o sinalefi e quella
della vocale iniziale della seconda parola ; quindi, se e ammissibile
che iva ajurydh possa aver dato iv’d°, non mi pare altrettanto probabile
che ririsista dyuh potesse dare ririsisf dyur ! Inoltre, il Bloomfield
puo si citare a p. 212 esempi in cui ris e costruito con dyuh, ma dyuh
significa di per se — e quindi anche in tutti i casi addotti dallo scienziato
americano — la vita „ , ma non gia “ la propria vita „ . E inline e
pur sempre da notare che I’accentuazione tramandata e ririflsUi yw
non °ld yur.
lo ritengo che la tradizione indigena non vada toccata ; quanto
al significato, non vi e dubbiochesi debba intendere “ danneggi quel-
Tuomo se stesso colle sue opere „ : bisognera quindi vedere in yur
una audace innovazione. E la cosa mi pare che si spieghi molto
facilraente. Se il poeta avesse avuto a sua disposizione due sillabe e
avesse potuto misurare lunga I’ultima sillaba di ririsista, egli avrebbe
detto ririfista tmdmrn jdnah. In assenza di questa possibility, egli e
ricorso ad un’ ardita formazione analogica ; poiche tmdn- pareva, pur
non essendolo in realta, una forma apocopata di atmdn- che significa
“ anima ,, e anche “ principio vitale „ (il significato “ ipse „ di atmdn-,
ereditato dal quasi omofono tmdn-, e secondario e seriore, vedi Rivista
degli Studi Orientali, xv, 364 sg.), il poeta si e creduto in diritto di rica-
vare da dyuh “ vita, forza vitale ,, uno yuh “ ipse La speculazione
grammaticale e antichissima in India, e noto come gia nell’ Aitareya
Brahmana venga frequentemente esercitata I’etimologia (cfr. Liebich,
Zur Einfuhrung in die indische einheimische Sprackwissenschaft, ii),
e casi come la creazione di un dhava- “ marito ,, in seguito a una falsa
divisione di vidhdvd “ vedova „ sono comuni a tutti i periodi dell’ ai., e
non solo di questa lingua. In yuh “ ipse ,, da dyuh secondo il rapporto di
trndn “ ipse ,, con atmdn- abbiamo una delle piii antiche testimonianze
del fenomeno in parola.
Neue Singhalesische Lautre^el
Von Wilhelm Printz
^X^ILHELM GEIGER stellt in Litteratur und Sprache der Singhalesen
(1900) § 23, 3 fest : ,, In einer Anzahl von Fallen ist der Palatal
c nicht zu s, sondern (dnrch j) zu d geworden." In der von Geiger
und Sir D. B. Jayatilaka unterzeichneten Einleitung zu A Dictionary
of the Sinhalese language (1935) wird das noch genauer bestimmt :
dies d entsteht nur aus intervokalischem *c, nie ini Anlaut. Zur
Erklarung heisst es ebenda we may assume that a later wav’e
of immigration brought to Ceylon a dialect in which c between vowels
was softened to j .. . The exact date of that immigration is un-
known. . . . “ Aber es ist doch sehr bedenklich, einen einzelnen
Lautwandel duxch Sprachmischung zu erklaren. Es lasst sich vielmehr
eine einfachere Deutung finden, die den allermeisten Fallen gerecht
wird.
Eine Musterung von Geiger’s Etymologic des Singhalesischen
(1897) ergibt 68 Worter, in denen singh. -s- oder -s auf alteres *-c-
zuriickgeht, wobei im Pali gewohnlich -cc- oder -cch-. seltener -iic-
oder -c- entspricht. Dagegen finden sich nur 11 Worter, in denen -d-
pali -c- entspricht. Hiervon bilden 9 eine Gruppe fiir sich ; adurd
(pa. dcdriya) ,, Lehrer “ ; kada (skr., pa. kdca, kdja) ,, Last “ ; godura
(skr., pa. gocara) ,, Beute, Nahrung “ ; narada (skr. ndrdcd, pa.
nardcd) ,, Pfeil “ usw. ; nidu (skr., pa. nica) ,, Mann ohne Kaste ;
niedrig “ ; padmn (skr. prdcina, pa. pdcina) ,, Osten wozu laut
Dictionary aaO. im 2. Jh.n.C. pajina ,, ostlich “ inschriftlich belegt
1 st ; yadinavd (ydc-) ,, bitten “ ; mudanavd (pa. mdceli ; nicht zu
muncati, wie Geiger angibt) ,, losen, befreien ‘b Fiir diese Gruppe
ergibt sich die Lautregel : hinter urspriinglichem Lang\mkal wird
*-c- iiber -j- zu -d-.
Hieran schliesst sich die Konjunktion da (skr., pa. ca), fiir die ja
die Zwischenform ja inschriftlich bezeugt ist. Wir miissen also
annehmen, dass bei dieseni enklitischen Wort die Stellung hinter
urspriinghch langvmkalischem Wortauslaut den Ausschlag gegeben hat.
Diese Lautregel gilt nicht, wenn der urspriingliche Endvokal
verstummt und c in den Auslaut gerat : pisas (pa. pisdca) „ Damon
Drei Worter fiigen sich dieser Lautregel nicht : vasa (skr. vac, pa.
vdcd) ,, Wort “ ; andrerseits mit -ad- aus -ac- : kavada (skr.. pa.
702
XEUE SIN'GHALESISCHE LAUTREGEE
kavaca) ,, Panzer ", woneben auch kavasa vorkoninit. sowie vuiana
(skr., pa. vacana) ,, Wort " ; vgl. ausser dein eben genaiintea vnsa
auch I'isi (pa. vaci) ,, Wort ".
Ausserdem gibt es noch vier Worter. in denen -d- nicht aus ein-
fachem -c- .stamrat und die daher gesondert zu betrachten sind :
ada (skr.. pa. anc-) ., gebeugt " ; {h)avitrudda (mit Metathesis aus
*avndum : skr. samvittsam. pa. sauimcchara) ,. Jahr ” ; duduru,
dxdulu (skr. dnsatra, im Pali nicht bclegt) .. schlechter Weg, Wildnis " ;
hihdu (falls mit sekundarer Xasalierung zu skr., pa. siici) ,. Stachel
(des Stachelschweins)
Griindliche Kenner des Singhale.si.schen werden wohl in der Lage
sein. weiteres Material herbeizuschaffen und die der aufgezeigten
Lautregel entgegenstehenden Falle besser zu deuten.
Les noms de la moutarde et du sesame
Par J. PRZYLrsKi et C. Regamey
A PROPOS du mot indieii sarsnpa, le Prof. J. Charpentier a emis
Fopinion suivante “ . . . Wolil am ehesten nichtarisches Wort ”d
Auparavant, le Prof. S. K. Chatterji avait deja ecrit : “ Skt. sarsapa
= Pkt. sdsava, which remains unexplained. But cf. Malay sesaivi.
(The Malay word may he a Prakrit borrowing ; but it is Skt., and not
Pkt., which furnishes Aryan loans in Indonesian.) ” -
On a en effet pour designer la moutarde :
malais sesaivi, sawi, sawi-sawi
javanais sesaivi
khmer sbey
bahnar habey, xabey
La plante qui produit la graine de moutarde, Brassica juncea ou
Sinapis juncea Linn, ou Sinapis patens Roxb., presente de grandes
analogies avec :
1) Brassica campestris Linn, qui est le colza indien ou sarson
(Sinapis glauca Roxb.).
2) Brassica oleracea Linn., qui est le chou.
Chou, rave, moutarde et colza indien appartiennent tous a I'espece
Brassica. II n’est done pas surprenant que plusieurs de ces plantes
soient designees par des noms analogue.s ou meme identiques. En fait,
voici quelques noms du chou de Chine, Brassica sinensis :
cam, siibei, bei
javanais, malais, said
batak, sabi
soundanais, sesaivi.
Ces noms sont apparemment des variantes d'un terrae generique
qu’on precise au moyen d'un second mot lorsqu'on veut specifier la
plante dont il s'agit. Ainsi sesaivi puteh d&igne en malais la moutarde
blanche qu’on appelle sesaivi pillau dans les parlers de la Peninsule
Malaise.®
1 Dans MO., 1932, p. 112.
“ Prearycin and Predrat'idtan in India, eel. par P. C. llagehi, Caleutta, 1929,
p. xxiv.
“ Blagden, Compar. VomhuL, s.v. mu.star<l.
704
J. PRZYLUSKI ET C. REGA>IEY
Tors ces noms se ramenent a une racine *safi a laquelle I'indonesieQ
adjoint le prefixe se-. Doit-on considerer ces mots comme empruntfe
a I'indo-aryen ? Cette hj'pothese se heurte aux difficultes suivantes ;
a) ]e mot ^tudie est un nom generique dans les langues austro-
asiatiques ; skt. sarsapa ne designe que la moutarde et c’est seulement
dans les langues indo-aryennes modernes que le mot signifie aussi
colza (sarson) ;
b) sarsapa est inexplicable par I’indo-europeen ;
c) les noms austroasiatiques different beaucoup de sarsapa ;
ils se ramenent tons a une forme avec i final sapi,^ et il leur manque
le prefixe sar-.
Au contraire, on explique aisement les faits si on considere sarsapa
comme emprunte a une ancienne langue austroasiatique. Le change-
ment de*-sapi en -sapa pent resulter de la tendance a integrer le nom
emprtmt^ dans la declinaison la plus usuelle. Aucune raison pbonetique
ou morphologique ne pent expliquer le passage de -sapa au *sapi
austroasiatique et il serait etrange que le mot eut toujours ete altere
de la meme maniere. L’insertion de r apres le prefixe sa- est conforme
a la morphologie austroasiatique oil les prefixes peuvent s’adjoindre
une nasale ou r. La difference entre les formes mon-kbmer actuelles
et le nom Sanskrit pent resulter de I'usure du prefixe ou de variations
dialectales. Le prefixe suivi de r etait assez frequent dans la langue
austroasiatique qui a fourni des emprunts a I'indo-aryen ; cf. skt.
karpdsa, karpata en face du khmer arntas, cam kapah, malais kapas ^ ;
skt. sarkara en face du stieng sbkar ® ; karpura en face du cam kapu,
khmer kapor ^ etc.
En khasi, le nom de la moutarde est tyrso et y note, dans cette
langue, la voyelle caracteristique des prefixes ; kyn-, pyn-, dyr-,
hyr-, etc.
Il parait done tres probable que I’indo-aryen sarsapa, forme d'une
racine sapa et du prefixe sar-, est un emprunt a une langue
austroasiatique.
L'etude du nom de la moutarde ramene I'attention sur un probleme
qui avait ete pose par Benfey des I’annee 1839 : “ . . . alvam, aLvanv.
^ Il est vrai qu’on trouve en Sanskrit le feminin sarsapi, mais ce mot signifie :
i) ein bestimmter Ausschlag, 2) eine Bachstelzenart {PW.y s.v. sarmpi), et ne peut
pas etre a la base des noms de la moutarde et du chou.
^ J. Przyluski, BSL., xxv, I, pp. 6&-71.
^ J. Przyluski, MSL., xxii, 5, pp. 208-210.
* J- Gonda, Austrisch en Arisch^ Utrecht, 1932, p. 23, et A. A. Fokker, Zeitsch.
fnr romaniaeh. Philol., 34 (1910), p. 567.
LES NOMS DE LA MOUTAEDE ET DU SESAME
705
In Skr. heisst er sarshajM und, wenn man annehmen darf, dass die
Griechen dieses, ohne alien Zweifel fremde, Wort von Persien her
erhalten haben, so ist eine Identitat von sarshapa und alvam nicht
gar u nm oglich. Denn die persischen Worte, welche r haben, erhalten
statt dessen im Pehlvi n ; so wiirde sarshapa — sansapa, woraus
durch Assimilation alvam werden konnte.” ^
Mais le mot ne parait pas atteste en iranien. D'autre part, les
faits grecs ne favorisent pas cette explication. L'ancienne forme
grecque etait ^ v&ttv (p. ex. Aristoph. Equ. 631 ; c’est aussi la seule
forme employee par Theophraste) ; cf. aussi vd-rreLov (Nic. Al. 430).
A I’epoque de la comedie moyenne apparait le verbe oLvam^ew
et ce n’est qu’a I’epoque hellenistique que alvaTrv, aivam est
atteste. Plus tard, la forme vd-irv est hors d’usage.
On trouve en latin des le debut (Plaute, Ennius) les formes sinapis,
senapis, mais on a en meme temps le mot napus “ chou-rave ’. Hehn
et Schrader sont d’avis qu’il est impossible de separer napus de vanv
et ces deux mots de aivanv. On se trouve done en presence d’un
doublet alvaiTv : vdirv dont I’origine n’est pas indo-europeenne.
Hehn ® dit a ce propos : " In den Gesetzen der Sprachen, aus der das
Wort entnommen wurde, konnte diese Doppelform begriindet sein,
aber welches war die Sprache ? Aegyptische Worter wie crtAt und
treat Al?, adpi (Aegyptische Wasserpflanze) und alaapov,
ferner Kopp-i, kIki., kv^i, dpp,i, orippi oder ori/Si, u.s.w.
lassen auch fiir vdrTv und trlvam auf aegyptische Herkunft raten.”
Mais I’explication suggeree par Hehn manque de base sohde. II
n’apporte aucim mot qu’on puisse considerer comme le modele
egyptien de vanv ou de alvanv et il ne prouve pas davantage que
ces plantes soient originaires de I’Egypte.
Si skt. sarsapa derive d'une racine anaryenne *sapi, on pent
assignor la meme origine a lat. sinapis, senapis et a gr. alvanv.
Les langues austroasiatiques sont caracterisees a la fois par 1' importance
des prefixes et par un systeme d’infixes dont I'un des plus frequents
est I'infixe nasal. On a par ex. en malais un homonyme de sawi
moutarde ” qui signifie “ de passagier die aan boord eenige
diensten verricht".^ Une autre forme de ce nom est senaui. II est
I Griechisches \V urzellerikon, Berlin, 1839, i, p. 428.
^ Cf. O. Schrader, Reallex. der Indgerm. AUertumskunde, 1901, p. 762 : Boissacq,
Diet. Etym. de la langne grecque, s.v. vdirv ; Walde, Lat. Etym. 11 art. p, 507.
^ Kulturpflanzen ®, p. 207.
* H. C. Klinkert, Maleisch-Eederlandsch Woordenboek, p. 358.
706
J. PEZYLUSKI ET C. BECAME Y —
clair que senawi derive de said par infixation de -di- et I on pent
former de la nieme maniere un derive *senapi a partir d’une racine
sapi. Des lors il n'y a plus rien d'obscur dans les formes latines et
grecques du noni de la moutarde. *Senapi est devenu en latin senapis,
sinapis. Le doublet grec vanv : mvanv est comparable au doublet
cam bei : snbSi ; la voyelle e de *senapi est tonibee et le groupe sn
s'est reduit a n. Plus tard a va-nv, emprunt iniparfait. s'est substitue
aiva-TTv, caique plus exact ou la voyelle e etait rendue par i conmie
en latin. ilais tandis qu'en grec vSittv et crivairv ont le meme sens,
lat. sinapis seul designe la moutarde ; napus signifie " cbou-rave
Le vocabulaire grec de I'epoque hellenistique ne conserve que la
forme aivanv, I'emprunt iniparfait vanv etant supprime. Par
contre. nupns s'est conserve en latin, mais a pris un sens un peu
different ; peut-etre a-t-il ete influence par le mot rapum " rave
II importe de noter que alvairv et sinapis sont beaucoup plus
proches des formes austroasiatiques que de sarsapa. Ceci semble
prouver que I'emprunt n’a pas ete fait par I'intermediaire de I'lnde,
mais qu'il result© d'un contact direct avec des gens parlant une
langue austroasiatique. La preuve de ces echanges a deja ete fournie
par 31. J. Gonda, qui a niontre que le nom latin de la cannelle,
cassia, est d'origine austroasiatique et ne se retrouve pas dans I'lnde.^
Skt. sarsapa, lat. senapis, sinapis, napus, grec vanv, aLva-rrv ne
sont pas des mots indo-europeens, mais des emprunts aux langues
anaryennes, car ils sont formes par des precedes des derivation
(prefixation, infixation) etrangers a I'indo-europeen et qui caracterisent
au contraire la farnille des langues austroasiatiques.
J. Cliarpentier - a suppose que le moyen-indien sdsavu est a la
base du nom grec du sesame ; arjaapov deja dans Alkman et Solon,
lac. aaaafioi’. aaapov. D’autres auteurs * croient plutot que le
mot grec est d'origine semitique, cf. arabe sdsim, simsim, aram.
sunisenui, susmd (ce dernier mot etant au.ssi a la base de remprunt
armenien s us may).
J. Cliarpentier incline a tirer du grec ou du moyen-indien la forme
arameenne et par suite le mot arabe plus recent. 3Iais cette hypothese
^ Tijd-srhr. A(inb\ Inst., Mei 193:2, pp. 717—727-
2 MO., 1932.
^ Cf. W. Muss-Amolt, “ WorcK in Greek and Latin,” Transact, of the Amer.
Philol. A.‘<s-or., vol. xxiii, 1892, p. Ill ; H. I.,ewy, Die Semitischen Fremdworter im
Oriechisrhen, Berlin, 189.J, p. 28.
LES NOMS DE LA MOUTARDE ET DU SESAME
707
est rendue peu vraisemblable par le fait qu'oii a en assyrieii
samaskimu}
On a essa3’e d'expliquer par le semitique les formes semitiques
du noni : F. E. C. Dietricb ^ fait deriver le nom arameen du sesame
de la racine signifiant entre autres sens : scharf sein. und stechen.
. . . Als Scbarfen werden nun auch die Gewiirze angesehen, pikant
schmeckende und so duftende : s^'r. ; a» ‘ cardamom ' ; gemein
semitiscbe sum ‘ Knoblauch ’ ; der duftendste, kostbarste Oel, chald.
sumsema ~ a-qaa/jLov." Fleischer® trouve cette etvmologie
insuffisante parce que le sfeame an und fiir sich in friscbem Zustande
geruchlos ist. " II explique le nom en partant de la racine smm " in der
der Begriff scbneller Beweglichkcit, geschiiftigen Hin- und Herlaufen
liegt '■ ce qui se rapporte, au point de ^me semantique, plutot a la
fourmi, dont le nom en arameen ressemble a celui du sesame :
sumsdna. kduidnd.
M. Jastrow * propose une autre etymologic : " aram. sumsum
(prob. from Mmes [sun-flower] sesamum (assyr. kimasktmu). Deriv.
snmkmd sesamum or poppy.”
Ces h^-potheses contradictoires ne permettent pas de considerer
comme acquise I'origine semitique de orjcrapov.
De Candolle est d'avis quo le sesame a ete introduit des iles de la
Sonde dans I’lnde ® et Watt rappelle que Sesamum indicum se trouve
a I'etat sauvage dans les montagnes de Java.® Jlais la question de
I'origine du sesame est obscure et ne parait pas pour le moment
susceptible d'etre tranchee d’une fa§on definitive.
Neanmoins il semble qu'on soit fonde a presenter les observations
suivantes.
1) Bien que le sesame et la moutarde ne soient pas de la nieme
famille, ils presentent certains caracteres communs. L'un et I'autre
sont remarquables par la petitesse de leurs graines ; celles-ci sont
oleagineuses et servent de medicament. Ces particularites pourraient
expliquer le transfer! du nom d une espece a I’autre.
2) Si Sesamum indicum est originaire des iles de la Sonde, il a pu,
en se repandant vers I’Ouest, recevoir dans le monde semitique un
nom caique sur celui de la moutarde. Il n'est pas impossible que ce
^ Code de Hammurabi, col. xiv, 22, 2.5, 31, 33, -47. 49. 59 ; col. xv, 3, etc.
- Ahhandlun-ge7i fur \Vortfor<r.hung, Leipzig, 1844, p. 64.
^ Dans I. Levy, Ckalddisches W drterbuchy 1867, ii, p. 578.
^ A Dictionary of the T argumimy etc., ii, s.v. sum-htm.
' Origine des plantes cultiieejf, p. 339.
® EcoiiomicaJ Products of India, p. 982.
708
LES NOMS DE LA MOUTARDE ET DU SESAME
nom se confonde avec celui que nous venons d’etudier. Les formes
semitiques du nom du sesame se ramenent a un schema consonantique
s{m)sm oh Ton peut reconnaitre un prefixe s(m) et une racine *sm
comparable a *sapi austroasiatique. Les formes comme saynassamu,
etc., seraient dues a une “ semitisation ” du mot etranger qui le
rapprochait du nom du soleil. Mais tout support chronologique fait
actuellement defaut pour asseoir cette hypothese et I’antiquite de
certains noms semitiques du sesame ne lui est pas favorable.
Dans ces conditions, il semble preferable de considerer au moins
provisoirement les noms de la moutarde et du sesame comme des mots
distincts et independants.
Sanskrit sd and sdh
By E. J. Rapson
^PHE forms of the nominative singular of the demonstrative pronoun
or definite article and of the relative pronoun in Sanskrit and
Greek correspond exactly in accordance with the regular sound-
changes of each language ; —
sd, sd, tad : 6, -q, t6 : I.E. *so, *sd, *tod ;
yah, yd, ydd : os, q, o : I.E. *ios, *id, *iod.
In Sanskrit the nom. sg. masc. has a double form — sd and sdh —
the uninflected base and the base with the regular termination of the
nom. sg. ; and in this respect Sanskrit finds its parallel in the Avestan
hd and has{cit).
So far, philologists seem not to have traced this characteristic
beyond the Indian and Iranian groups of the Indo-European family
of languages. Thumb, for instance (Handbuch des Sanskrit, § 361)
suggests that the forms with the nom. sg. termination may represent
eine arische Ncubildung. There can, however, be little doubt that the
Greek counterpart of sdh is to be seen in the Homeric demonstrative
pronoun os ; e.g. dAAd koI os SeiSot/ce “ even he fears ” {Iliad,
xxi, 198). This demonstrative os is very rare in later Greek, but it
survives in certain stock phrases : e.g. q 8'os, “ said he ".
We find, then, in early Greek no fewer than three different words
which are all of them regularly represented by os — the relative
pronoun, Skt. yah, I.E. *ios ; the demonstrative pronoun, Skt. sdh,
I.E. *sos ; and the possessive pronoun, Skt. svdh, I.E. *suos.
It is commonly assumed in Greek grammars that os “ he ” is
simply the relative used as a demonstrative ; and this ancient belief,
dating from a period long before the existence of comparative
philology, has become traditional, and has prevented Greek
grammarians from seeing that this os and d are inseparably connected
not only in meaning but also in origin.
Indian grammarians on the other hand have declined to admit
that sa has an independent existence of its own : for them sa is
merely a truncated form — sah with lopa of su, the termination of the
nom. sg. masc. Thus in the Pada Patha of the Rigveda, which repre-
sents the first step in the history of Sanskrit grammar — the rydkarana
of a sentence into the words of which it is composed, and of certain
710
E. J. EAPSON
compounds into their constituent elements — tlie sa of the Sainhita
Patha, in whatsoever connection it may occur — whether before a
consonant or before a vowel or as combining with a following vowel
to form a diphthong — is invariably restored as mh.
sn devum ehn vaksali (I, i, 2).
Pada : sdh-devan.
sd id devesu gacchati (I, i, 4).
Pada ; snh-it.
yam smd prcchdnti Jcnha seti ghordm (II, xii, 5).
Pada ; sdh-ili.
The Indian grammarians were in fact wedded to a theory from
the very beginning. They started with a postulate ; ev'ery inflected
word must have its proper suffix. If, then, the suffix is not manifest,
its absence is due to adarsaiia and it must be supposed to exist. This
is the doctrine of lopa or “ omission ”, which Western grammars
of Sanskrit have inherited from the Indian grammarians. Thus
Whitney’s statement ;
“ The nominative ma.sculine pronouns sds and esds and (Vedic)
syds lose their s before any consonant (A San.-ikrii Grammar,
§ 176a) is simply a paraphrase of Panini, VI, i, 132-3. Panini
goes on to say (VI, i, 134) that for metrical reasons the same
elision may take place before a vowel (with the consequential sandhi
of the two vowels) ; and the Rigveda Pratisakhya gives two lists (172
and 173 ; ed. Max Jliiller, pp. 54-5) comprising twenty-six illustrations
of this fusion of sd with different vowels ; e.g. sdsmin for sd astnm ;
sendra for sd indra ; sese for sd Ise ; sopama for sd upamd ; saind
for sd end ; saumdhlh for sd osadfdh. In all these and similar instances
sd i.s represented by sdh in the Pada text. But, as a matter of simple
observ'ation, these “ exceptions ” are so frequent that they supersede
the rule. They are in fact in accordance with the general rule :
” sd in RV. is in the great majority of cases combined with tl.e
following vowel ’’ (Whitney, § 1766).
The facts of the Vedic language are thus in agreement wdth the
evidence supplied by comparative grammar, viz. that, like ha and
hasicit) in Avestaii and d and d? in Greek, sd and sdh are alternative
and independent forms of the demonstrative pronoun. In the Veda
there is no such rigid distinction observable in the employment of the
two forms as is ordained by the grammarians for Classical Sanskrit :
so IS no doubt normally used before a consonant, but it is also
711
SANSKRIT SA AJJD SAH
frequently used before a vowel ; while sdk is always used in paitsa
and normally before a vowel. But there are a few undoubted instances
of the occurrence of sdh followed by a consonant, particularly by
t and p : —
naJii ms tdva no mama
sdstre anydsya rdnyati, VIII, xx.xiii, IG.
yo no dvesty ddharait sds padlstu. III, liii, 21.
When at a later date grammarians came to consider the prevailing
usage, they laid down hard and fast laws : sa must always be used
before consonants, and salt before vowels and in pausa. Piinini
assumes that before vowels the sandhi of the -ah in sah is precisely the
same as that of any other -ah — so 'brarlt, naro 'hravlt (P., ^’I, i, 109) ;
sa cti, nara eti (P.. VIII, iii, 18 and 19) ; that is to say, Panini holds
(1) that -ah before all voiced sounds, whether vowels or consonants,
becomes -o ; (2) that this -o absorbs a following initial a - ; and (3)
that -0 becomes -av before any other vowel. So far therefore, except
as regards (2) in word-formation, e.g. bhavati, the principles of external
and internal sandhi in Classical Sanskrit are identical ; and as regards
(3) it is explained that, when -av comes at the end of a word, the semi-
vowel -V being very lightly pronounced {laghuprayat nalara) tends to
disappear : in the opinion of Sakalya it does in fact disappear ; in
the opinion of other grammarians its retention is optional, i.e. either
visnav ehi or idsna ehi is allowable.
Curiously enough. Western scholars in their explanation of this
sandhi have sometimes adopted the doctrine of lopa and extended it
beyond the limits contemplated by Piinini. Thus Whitney says
(§ 175c) : ■“ final as before any other vowel than a loses its s, becoming
simple a. "
There can be no doubt that, in this respect at least. Panini has shown
us the better way.
Pali bhunaha
By Baburam Saksena.
ri'^HE word bhunaha occurs thrice * in Canonical Pali literature :
J- (a) in the Sutta-nipata ® as sg. voc. ; (6) in the Majjhima
Nikaya ® as sg. gen. bhunahuno ; and (c) in the Jataka * as pi, nom.
bhunahimo.
The Sn. commentary (p. 479) explains bhunahu as bhutihanaha,
vuddhi-nasaha and the Jataka commentary (as quoted in the Pali
Diet.) interprets bhunahuno as mnam ativattaro attano vaddhiyd
hatattd bhunahuno. Lord Chalmers in his translation of the Majjhima
Nikaya suggests ” puritanical ” as the sense of bhunahuno, while
Rahula Sahkrtyayana in his Hindi translation ^ appears to have taken
the word as a proper name (bhunabhu), used as an adjective of seyyam
and not of Gotamassa ; he leaves the term unexplained even in the
glossary.
Dr. \V. Stede, in his Dictionary, notes the difficulty in the
explanation. of the word, and asks: ‘‘Is it an old mis-spelling for
bJmta -f gha ? The latter of han ? " Then he suggests “ a destroyer
of beings ”. One does not find the word in Childers's Dictionary. Bapat,
in his edition of the Sn. (Poona, 1924, p. 181), merely quotes the com-
mentary to explain the word.
Pali scholars have been misled by the commentaries in the inter-
pretation of this word, bhunaha most certainly corresponds to Skt.
bhrunahan-, bhrunaghna- “ the killer of an embryo, one who produces
abortion The word bhruna- occurs as early as the Rigveda (x, 155,
2).® The production of abortion is considered a heinous crime in the
Mahabharata ; the Manusmrti prescribes a very heavy punishment
for it. It was then, as even now, one of the most condemned crimes
^ Rhys Davids and ^^tedc : Piili Dictionary, vol. iii, p. 132.
^ verse 664 : inukhadugga vi^huta-m-anariya
bkumhii papaka dukkatakari.
vv. 11. hhiuiahata, bhunahota, hhuhata.
3 dufldittham mia bho Bhiradtdja addasama ye mayam ta-ssa bhoto Gotamassa
bhunahiuio ■■^eyyam addasamati.
* ed. Fausboll vol. v, pp. 266, 272.
^ Published bv the Mahabodhi Society. Sarnath, Benares (1933), p. 292 : bho
Bhnradvdja ! yah burd dikhnd hud, jo ham ne dp {ke?) Gautam kl Bhunabhu, sayyd
kd dekhd.
^ M. Williams, Skt. Eng. Diet., bhruna.
VoL. VIII. PARTS 2 AND 3.
46
714
PALI BBVSABA
in India. I quote the Vacaspatya (p. 4711) for the significance of the
word hhrunaghna : — bhrunaghna — tri. bhrunatn garbham hanti, han
-ka. bhrunahatya-kdrake kvip, bhrutiaMpg atra ; “ ajpi bhrunahanam
nidsdt.” — Manuh.
With this interpretation of the word we arrive at very suitable
sense for the passages where it occurs. In the Sn. the word
stands on a par with terms indicating “ ignoble, sinner The
Maj. Ni. has the passage as an accusation by Magandiya, a heretic
Brahmin, against the Buddha, and if the word did not contain an
abuse of the Lord there would be no point in Bharadvaja, the host
of the Lord, remonstrating with Magandiya and asking him to with-
draw the accusation. As such Chalmers’s suggestion is unacceptable
and Rahula Sankrtyayana's interpretation evidently meaningless.
It may be pointed out that this is not the only place where the
Pali commentators have failed to catch the correct meaning due to
their ignorance of Sanskrit and to their unfamiliarity with the cognate
Sanskrit literature. As I have shown elsewhere,^ it is not only the Pah
commentators, but sometimes the redactors of the canonical works
also who offer fantastic and uncalled-for etymologies.
^ Fantastic Etymologies in the Dhammapada, Ganganatha Jha Commemoration
Volume.
Historia de Gramatica Concani
By Martaxo Saldaxha
. desde tempo remote passa por verdade demonstrada o
grande absurdo de que a lingua desta terra (Goa), a lingua de meio
milhao de bomens, nem tern gramatica nem e susceptivel de ser escrita ;
julgamos que ao menos faremos algum servigo se dissiparmos tao
perniciosa ilusao, mostrando que a lingua concani nao so tern a sua
gramatica, como qualquer outra, mas que a mesma gramatica foi
em tempo formulada em regras e ate impressa.” J. H. da Cunha
Rivara, Ensaio Hislorico da Lingua Concani, 1857, pg. 1.
^\RIGEM E VALOR DAS PRiMEiRAS GRAMATiCAS : 0 concani e a
lingua vernacula de Goa (India Portuguesa) e dos paises circun-
vizinhos, sendo hoje falada por inais de uni milhao e meio de pessoas ;
e. apesar de ser muito propagada a instrugao na sua area, e talvez
a unica lingua indiana, que, alein de nao ter cultura literaria, nem
mesmo e objecto de ensino, oficial ou particular, no seu pais,^ cujos
filhos preferem para esse fim ou linguas extrangeiras, como portugues
e ingles, ou as vizinhas marata e canares. £ verdade que, com
0 advento da civilizagao luso-crista, os antigos missionarios, desde
o seculo XVI, ministraram o ensino desta lingua, mas foi somente
aos seus eolegas europeus para os fins de catequese,^ e nao aos naturais,
para o que compuseram gramaticas e vocabularies, que seriani minas
preciosas para a filologia Indiana,^ se tivessem todos chegado ate nos.
Deste modo o concani, dentre todos os vernaculos modernos da India,
foi o primeiro a possuir uma gramatica, gragas aos esforgos do
benemerito jesuita ingles Tomas Estevao.
Esta primeira tentativa, alem de facilitar aos extranhos o estudo
da lingua, estimulou e ajudou a organizagao de novos trabalhos
‘ Urn professor de instruyao primaria (Pe. Apuleio da Cunha), depois de aposentado,
dedieou-se ao ensino particular gratuito da lingua na Provincia de Pernem, obtendo
resultado animador. JIas a eseola, que era movel. morreu com ele. Ultimamente
foi introduzido o seu ensino na Eseola Superior Colonial de Lisboa.
- “ . . . desejando (Fr. Cristovam de .Jesu.s) instruir na lingua canarina aos seus
companheiros para o fim de conduzireni ao gremio da Igreja a muitos barbaros,
escreveo : Arte Grammatical da Lingoa Canaiina . . . " Vergel das Plantat, de Fr.
Jacinto de Deos. cap. i, pg. 10.
^ “ It (konkani) also has a large literature mostly dating from the glorious times
of the early Portuguese rule at Goa, and due to the surprising zeal and ability of the
former Jesuits, and for these reasons must be of great interest to a philologist.” A C
Burnell. Specimens of South Indian Dialects, n.° 1 (konkani).
716
M. SALDANHA —
similares devidos todos, ao principio, a penas estrangeiras, aparecendo
so mais tarde, desde o seculo xix. obras congeneres de estudiosos
natives. E assira, o numero das gramaticas concanis, impressas e
manuscritas, ate hoje conhecidas. sobe a duas diizias ; acham-se
escritas em portugues, espanhol, latim, frances, ingles, marata. concani
e talvez era canares, algumas das quais so se encontrara hoje nas
bibliotecas eruditas de Londres. A pobresa dos missionarios concorreu
para que algumas dessas obras nunca logra.ssem ser impressas.^
A inaior parte de.stas gramaticas, tendo sido escritas em paises
e epocas diferentes, representam as formas coloquiais correntes, nos
liltimos cinco seculos, desde Ba^aim e a Ilha de Salsete no norte,
ate Mangalore no sul ; e, como tal, a sua leitura e muito proveitosa
para o estudo comparative e evolutive da lingua nos seus variados
dialectos, mesmo admitindo que sejam inexactas em algumas
afirnia9oes, o quo, alias, nao e de extranhar por parte da maioria
desses autores, para quem o idioma nao era vernaculo, quando se
repara que algums dos liltimos grainaticos. nao obstante serein filhos
do pais, nao teem escapade a incorrec96es.
A presente memoria. porem, nao visa a este e.studo comparative,
ou a gramatica historica da lingua, mas unicamente a historia da sm
gramntica, i.sto e, tenta orgauizar dessas gramaticas uma rela9ao
bibliografica completa, quanto permitem os conhecimentos actuais,
e indicar a maneira como as mais importantes dentre clas interpretaram
e resolveram os principals problenias gramaticais.
Tratando-se de bibliografia concani, nao se pode deixar em silencio
o venerando nome de Joarjuim Heliodoro da Cunha Rivara, a quem
se deve a primeira e linica bibliografia geral. incluindo a gramatical,
embora incompleta, publieada no .sen valioso Eamio Hiatorico da Lingua
Concani, que ainda lioje nao perdeua sua actualidade. £ste inesquecivel
concanofilo. tendo ido para a India em 1856, como Secretario Geral
do Governo e Comissario dos Estudos, empregou, infelizmente sem
proveito, o seu melhor esf6r90 pcla cultura da lingua, quer
recomendando o seu ensino e dando as necessarias instru9oes, quer
iniciando, logo em 1857, a publica9ao de tres gramaticas e um
dicionario, os quais todos, com a excep9ao da gramatica de Estevao,
^ “ Xao se tern dado ao prelo muitos destes livros visto carecermos de cabedais
pela pobresa evangelica que profe^aiiios e os que se deram a estampa se devc a
diligencia dos autores que souberam dedicar os seus Iivros a pessoas Reais e lilustres,
as quais mandaram imprimir com seu dispendio. " IlisUtr. dos (Jonv, e Coh/J., e/B., de
S. Thome da Provmeia de 8. Francisco em Goa, pg. 31 ; Ms. Bib!it. Xal. F.G., 177,
pg. 31.
HISTORIA DE GEAMATICA COXCANI
717
se encontravam em Ms. e se teriain certamente perdido para a
posteridade, se ele os nao tivesse salvado da destrui§ao, criando assim
para a sua memoria jiis ao eterno reconhecimento dos povos concanis.
0 NOME DA lingua : Antes do mais, porem, e necessario desfazer
uma confusao em que pode trope^ar o leitor por causa da variedade
de nomes com que e designada a lingua, tais como canarim, concanim
gomantaJci, brdmana, goani, etc.
Como a primitiva lingua dos indo-arias, que nao tinlia nome e.special,
sendo mais tarde conhecida por sanscrito, assim o idioma goes, que
e irmao do marata, era apenas designado pelos naturais como dm chi
bhds, a nossa lingua e fora do pais por govi ou Goenchi bhds, lingua de
Goa. Entre os maratas eruditos e geralmente conhecida por gonmntaM,
de Gomantaka, nome sanscritico de Goa.
Os antigos portugueses, incluindo os missionarios, deram-lhe
varios nomes, qual deles mais desarrazoado, sendo o mais vulgar o de
“ lingua canarim ’’ , um absurdo que, talvez por o ser, mais se apegou
e generalizou. Niuguem ignora que o idioma goes e indo-arico ou
sanscritico, irmao de marata, nada tendo portanto de comum com o
canari, cuja gramatica e o lexico sao muito diferentes, como lingua
dravldica que e, e alheia ao sanscrito. Os missionarios, que estudavam
os vernaculos e muitos andaram pelo Canard, deviara reconhecer a
impropriedade do termo ; mas, como nao escreviam para filologos,
seguiram a pratica corrente, Assim, Estevao designou-a por “ lingua
canarim " e o missionario portugues (infra n.° 2), cuja gramatica,
alias, trata do marata concanico das pro\’incias do norte, Bafaim,
Bandora, Bombaim, chamou-a tambem “ Arte Canarina da lingua do
Norte Todos os autores, porem, reconhecem em Goa duas formas
de falar : a plebeia, um tanto irregular, chamada canarim e a outra
mais regular, usada pelas classes cultas, a qual chamavam lingua
canarim brdmana ou simplesmente brdmana de Goa. Como era esta
ultima variedade a preferida pelos europeus, e mesmo por outras
castas, para a escrita, sermoes e uso religioso, foi ela tomada para
norma por todas as gramaticas, incluindo a de Estevao, a qual, apezar
de se dizer da lingua canarim sem acrescentar brdinana, cingiu-se,
contudo, ao falar bramanico, como se ve do seu contexto ; tambem
a licen9a do ordinario para a sua impressao se Ihe refere como “ arte
da lingua canarin bramana ”, Para o seu Parana, porem, escrito para
a leitura dos nativos, Estevao preferiu. como mais apropriada para
a nobresa do assunto, o marata, Mardttiye bhdssena lihile dhe, como
diz, embora substituisse algumas formas e locu^oes eruditas por outras
718
M. SALDANHA
de linguage bramana ’’ local (concani), como niais compreensiveis
do vulgo. Os missionarios que foram encarregados de dar o seu parecer
para a primeira edi 9 ao deste Purana dizem-no escrito em “ linguage
bramana marastta ”, distinguindo-a assim da vernacula, chamada
bramana canarim ” e nao marastta. Fazia-se, portanto, ja nessa
epoca distin^ao entre o chamado canarim on dialecto de Goa e o
marata, on marastta (de Maharastra).
Os primeiros gramaticos que empregaram o termo proprio foram
Arcamone (seculo xvii, n.° 13) e o missionario italiano (sec. xviii,
n.° 7). Arcamone chama “ lingua concanica " a de Goa e “ decanica ”
a marata ; e o italiano chama- a “lingua Concana ” e “ de Concana ”
e as vezes tambem “ concani ” ; mas esta ultima palavra pode bem
ser modificagao de Eivara, analoga a que fez na impressao doutras
gramaticas, substituindo Arte Canarina por Gramdtica Concani,
para evitar confusao. No seculo xix aparece tambem empregado por
alguns cristaos de Goa o nome Gomantaki, usado em marata.
Preferiram-no ao concani, tanto para destinguir o idioma goes
do concani falado no norte do Concao, como para ligar o nome da
lingua ao do pais (Gomantaka = Goa). Actualmente a designa 9 ao
mais adoptada e a de concani. Houve, porem, um gramatico moderno,
que, inspirado talvez pelo ingles goanese, preferiu chamar-lhe goani,
nome inteiramente anti-etimologico e exotico, que o proprio autor
abandonou, em vista de razoes que Ihe foram expostas, mas nao sem
nos informar que “ alacremente fora aceito por muitos ” talvez para
justificar os psicologistas, para quern o povo sso tern alacridade para
aceitar o que for errado, como canari em vez de concani. 0 unico
nome que seria mais apropriado ao dialecto de Goa, como sendo ao
mesrao tempo popular, geografico e etimologico, e o que corre entre
os povos circunvizinhos, que Ihe chamam govi (de Govd ou Govern
= Goa), como em portugues seria goes. Costuma-se ainda designa-lo
em portugues local por lingua da terra ou simplesmente por lingua,
como em ; F. pregou em lingua, i.e., em concani.
Bibliografia ; 0 primeiro autor a figurar nesta bibliografia e, como
ja se disse, o jesuita ingles Tomas Estevao (Thomas Stephens), cuja
gramatica e o trabalho pioneiro no genero.
Nascido em Inglaterra em 1549, Estevao partiu de Roma para
Lisboa e daqxii para a India, tendo vivido em Goa desde 1579 a 1619,
em que morreu com a idade de 70 anos no colegio de Rachol, de que
era Reitor. A Estevao cabe a primazia em alguns factos de historia
e linguistica indianas. Pois foi ele o primeiro ingles que chegou a
Bull. 8.0. S. Vol. VIII, Parts 2 and :5.
Plate I\'.
““I lESVSMA^lA
ARTE .D A
!,L I N GOACANA
l^j RIM
p COMPOST A pELO PAD RE.
pq Thoibaa eftcuaoda Conipanhia d««-
IISVSS^ acrcccntada pcllo padr-
Diogo Ribciro d^ niefma CSpanhia
^ E noucQitnc€ reuIfta.fiCcmcndada por
outros tjuairo Padres da Com
panhisjt^.
MS
I
pTc^ Licenca da S. Inqolficam 5C Or
^ dtiiario
^^jem Racfiol no Collcgio dc S. Ignacio
§ da Compaahia dc lES V' Anno dc
I ■.
y- 71 .
nMilCTg^ |g>^ pv’Og^ccig^cgC!:.g^a5WQ»
HISTORIA DE GEAMATICA CONCANI
719
India via Cabo e foram as suas cartas escritas ao pai, quo, segundo
dizem, animaram os ingleses a tentar as prinieiras rela 96 es comerciais
com a India ; e em Goa empregou a sua influencia em favor de sens
patriciosd Foi ele o primeiro europeu que organizou unia gramatica
indiana (a de concani) ; e foi tambem queni compoz o primeiro
catecismo na mesma lingua ^ e o primeiro poema cristao em marata
(o Purana da bistoria biblica) ® ao qual cabe a gloria de ser a primeira
obra em lingua indiana, de carater literario, impressa (1615) em
carateres romanos ; e o seu prefacio e citado como um dos primeiros
especimens de prosa marata “ ; e, finalmente, foi ainda esse mesmo
ingles, que, dois seculos antes do seu patricio William Jones, comunicou
para a Europa a semelhan 9 a entre as linguas indianas e o grego e o
latim. “ Many are the languages of these places — escrevda ele para
seu irmao em outubro de 1583. — Their pronunciation is not disagreable
and their structure is allied to Greek and Latin. The phrases and
constructions are of a wonderful kind. " ®
0 nome de Tomas Estevao hem merece, portanto. nao so da Igreja,
mas tambem de Portugal, Inglaterra e India, e especialmente dos
idiomas marata e concani, cujas cristandades, em algumas partes, se
deleitam ainda hoje com a leitura do seu Purann.^
A gramatica de Estevam foi impressa com estes dizeres :
1 — lESVS MARIA I ARTE DA | LINGOA CANA \ RIM,
composta peh Padre \ Thomaz estevao da Cotnpanhia de \ lESUS <£•
acrecentada pello Padre | Diogo Ribeiro da mesma Copanhia. | E
nouamente reuista, e emendada por | outros qimtro Padres da mesma
Companhia : jl^^S ( Com Licen^ da S. Inquisigam tfe Or ) dinario, |
' Eram Ralph Fitch, e mais trcs companheiros, que, tentando uma viagem
terrestre para a India, foram presos pelos portugueses em Ormus, e mandodos
para Goa.
^ Intitula-se : “ Boutrina Christa Em Lingua Bramana Canarim Ordenada a
maneira de Dialogo pera ensmar os mininos. Coposta polio Padre Thomas Estevao da
Companhia de lesus, natural de Lodres, 1622.” O de S. Francisco Xavier, que o
precedeu, era cm portugues.
^ Editado pela 4a. vez por Joseph L. Saldanha com uma valiosa Introdii<;ao,
Mangalore 1907. A obra tem imitado tao bem o estilo e a linguagem dos puranas
hindus, que um autor de litcratura marata (Bhave) supoe seja cscrita por algum hmdu
sob a direc^ao de Estevam.
* Vid. R. Bh. Joshi, Marathi Bhaxechi Ghatna, pg. 304 e seg.
“ Vid. 0 Purana. Introdu^ao de Saldanha.
® Os 60 mil prisioneiros de Mangalore, levados por Tipu Sultao para Seringapatan,
confortavam-se, nas agruras do cativeiro, lendoem comum fete Purana ; cit. Saldanha.
Xas provincias do Xorte e mais lido o Purana de Francisco Vaz Guimaraes,
transliterado pelos natives em carateres maratas.
720
M. SALDAXHA —
em Radiol, no Collegia de S. Ignacio \ da Companhia de lESU. Anno
de I 1640 (Vid. Estampa)d
Foi pois impressa 21 anos depois da morte do autor, sendo hoje
impossivel saber-se qual a parte original e quais os acrescentamentos
e emendas dos outros padres. Desta edi§ao supunha-se existirem so
dois exemplares e ambos em Londres,^ pertencentes um a Biblioteca
do India Office, que, porem, ja o nao possui, e o outro a Biblioteca da
School of Oriental Studies, o qual ainda la existe, faltando-lhe a
primeira pagina e algumas no fim. A Biblioteca Nacional de Lisboa,
porem, possui um exemplar inteiro e bem conservado. 0 frontispicio
e encerrado em moldura grafica. Devem ser, portanto, estes dois
liltimos os linicos exemplares que, actualmente, se sabe existirem.
0 livro foi reimpresso por Cunha Rivara sob este titulo ;
■■ Grammatica da Lingua Concani, composta pelo Padre Thomaz
Estevao e acrescentada por outros Padres da Companhia de Jesus.
Segunda impressao correcta e annotada ; a que precede como
introdu^ao a Memoria sobre a distribuifdo geografica das principais
linguas da India, por Sir Erskine Perry, e o Ensaio Historico da Lingua
Concani pelo Editor. Nova-Goa ; na Imprensa Nacional 1857. ’
Esta primeira gramatica e a do n.* 7 representam a variedade coloquial
de Salsete.
2 — Arte Canarina na Lingoa do Norte. Ms. anonimo, que
“ pelos seus carateres extn'nsecos e di^ao ” conjecturou C. Rivara ser
da autoria de " algum Religioso Franciscano ou da Companhia de
Jesus, residente em Thana, na Ilha de Salsete, onde, sem diivida, foi
composta a obra E, por isso, ao edita-la, deu-lhe esta fachada :
■■ Gramatica da Lingua Concani no dialecto do Norte, composta no
seculo XVII por um Missionario Portugues ; e agora pela primeira
vez dada a estampa. Nova Goa ; na Imprensa Nacional ; 1858.”
E por isso conhecida como a “ gramatica do Missionario, Portugues ”.
1 Dando o seu pareoer para e.sta impressao, diz o P.® Estevao da Cruz : “ achei
a lingua toda reduzida a regras e pre^-eitos de grammatica muito ^ertos, e bom e facil
estilo pera se poder aprender com faciiidade . . , e ob autores todos della de muito
grande louv'or." Estc P.® Cruz c o mesmo que compos em linguagem bramana
marastta um poema intitulado Dihcurso sobre a vinda do Apostolo S. Pedro ”
ao qual Rivara chama 0 Purana da Biblioteca (de Goa) por ignorar o nome do autor
pela razao de estar truncado o re&pectivo exemplar. Ele nao era portugues, como
vsuj)6s Barb. Machado, mas sini frances, como se le no mesmo Discurso. Curioso e
que, como se deseonhecesse o Purann de T. Estevao, diz que RvS materias do om poema
^am muitas delas pouco ou nunca tratadns nestt estylo e. Uttguagem e assim ndo poudenws
ter a quern imitar.
- Cit. Saldanha, pg. 37.
HISTORIA DE GRAMATICA CONCANI
721
Esta e a primeira gramatica iiiarata, escrita em lingua europeia
e representa o dialecto concanico do norte, ou o niarata coloquial de
Ba 9 aim, Bandora, Bombaim, e outros territorios do norte do Concao,
ao tempo sob o dominio portugues, do mesmo modo como a do n.° 16
representa o marata decanico. Embora nao digain respeito propria-
mente a lingua goesa, nierecem estas ambas ser aqui inencionadas
como um bom auxiliar para o estudo comparativo concano-marata.
Bern diz Rivara : " Pareceu-nos que esta (do n.^ 2), alem de ser um
novo monumento dos trabalbos literarios dos portugueses no Oriente,
seria um complemento nao so util mas indispensavel a Gramatica do
Padre T. Estevao ; pois, sendo ordenada pelo mesmo piano e sistema,
facilmente se confrontarao, por meio dela, as diferen^as dos dialectos
concanis de Goa e Ba§aim e se elucidarao e confirmarao mutuamente
as regras e idiotismos da lingua geral. ”
provavelmente a esta gramatica, ou a do n.° 16, impressa em 1778.
que se refere Carey na sua " Grammar of the Mahratta Language, by
W. Carey, teacher of Sanscrit, bengalee and mahratta ... in the
College of Fort William ; Serampore, 1805. " Diz Carey : "A grammar
of this language was indeed written many years ago in the Portuguese
tongue ; but the writer of this not having been able to procure a copy
of it, could not derive any assistance from the labours of its author,
and has therefore been obliged to strike out a plan of his own." E
foi pena, por que a gramatica portuguesa teria fornecido a Carey
muitos elementos que Ihe escaparam.
3 — “ Arte da Lingua Canarina, por Fr. Gaspar de S. Miguel,
4°, Ms.” F assim mencionada por Barbosa Machado na sua Biblioteca
Lusitana, donde a cita C. Rivara, que nao a viu. Talvez seja a mesma
obra cuja copia existe na Biblioteca da Sch. Or. Stud. (Mar.sden
Collection, ii, 559, n.° 1) e e assim intitulada : " lesu Maria loseph.
Gratmnatica da lingua bramana que corre na Ilha de Goa e sua comarca.”
Ms anonimo, seguido imediataniente, no mesmo Codex, doutro
(n.° ii), que e : " Sgntaxis copiosissima na lingua bramana e pollida
composta pello Pe. Fr. Gaspar de 8. Miguel, Portugues, frade menor,
Pregador e mestre. dela etc." (era friide de S. Tome da Pro\uncia de S.
Francisco, em Goa). Sao 39 folhas ou 78 paginas. contendo 205 regras.
Pelo desenvolvimento da sintaxe, que esta em propor 9 ao com o do
n.° 1 (fonetica e morfologia), parece que os dois niimeros sao partes da
mesma obra — a gramatica de Fr. Gaspar.
4 — “ Arte Gramatical da Lingua Canarina, por Fr. Christovam
de Jesus ; Ms.” Assim citada por C. Rivara, tambem sob a autoridade
722
M. SALDANHA
de Barbosa. Mas a copia que se encontra na mesma School of Or. St.
(Marsden Col., ii, 559, n.° iii) traz este titulo : “ Gramatica da Lingua
Bramana ordenada | pel-lo Pe. Fr. Christovao de lesus no anno de (
16.35.’’ Tem 42 (?) folhas, em letra muito fina, mas bem perceptivel.
Come^a pelas declina96es.
A impressao destas duas ultimas gramaticas seria interessante para
o estudo da variedade coloquial de Bardes, onde missionava a ordem
dos franciscanos, a que pertenciam os dois autores.
5 — “ Arte de Gramatica da Lingoa Bracmana, disposta em II
Livros ; Obra mui necessaria para os Missionarios, Pregadores,
Confessores, Compositores, Poetas e Estudantes nas partes deste
Oriente. Na Ilha de Chorao. Escrita por Estudante Simao Al'z
(Alvarez) Bragmane, Semnoye, Choranense, etc. Anno mdciciv. ’’
E Ms., de autor desconhecido, sendo talvez o Semnoye Choranense,
como entende Rivara, um simples copista, devendo por isso a palavra
*■ escrita ” ser entendida por tresladada.
6 — Gramatica citada (sem titulo) por Francisco Luis Gomes
(n.° 10) como “ ante-posta ao Ms. de um Dicionario escrito no ano
de 1695 ”. Provavelniente e a gramatica n.° 5, que e acompanhada
de um “ vocabulario em tres linguas, escrito por Simao Al’z e seu
pai Lourenfo Al’z, mdcicv”. Pelas poucas referencias que Gomes
Ihe faz parece ser um trabalho interessante e e pena nao ter sido
impresso.
7 — Grammatica oh Observances Grammaticais sbbre a lAngoa de
Concand. Ms. anonimo, impresso por C. Rivara sob o titulo :
“ Grammatica da Lingua Concani escrita em Portugues por um
Missionario Italiano. Nova Goa na Imprensa Nacional ; 1859.” “
escreve Rivara, segundo todas as noticias que temos podido obter,
obra do Padre Fr. Francisco Xavier de Santana, italiano, Carmelita
descal50, Missionario no Canara, Arcebispo de Sardes e ultimamente
\ igario Apostolico em Yerapoly. Da obra se ve que foi escrita no
Canara e na lingua portuguesa com seus resaibos da italiana, os quais
tomamos a liberdade de corrigir, sem contudo alterar no minimo o
pensamento do autor. ’ de extranhar, porem, que este erudito
editor, que, alias, a proposito do Ms. n.” 2, indica o seculo em
que o supoe escrito. nada diga neste ponto com respeito a esta
gramatica, nem mencione a data da estada na India de Fr. Francisco
Xavier, limitando-se a dizer “ ultimamente Vigario Apostolico ’
mantendo a mesma atitude todas as vezes que se refere a esse frade
(cf. Ensaio Hist, da Ling. Cone., pg. xxxix n.b e pg. cxviii ; e Dicion.
HISTORIA DE GRAMATICA CONCANI
723
Cone. Port, composto por um Miss. Ital. 1869, Pref. pg. i). A sua opiniao,
porem, da margem a certos reparos ;
I Nao ha na obra cousa que denuncie ter sido escrita no Canard,
a nao ser um parentese nesta regra da pg. 24 : “0 modo tinico de
fazer fraegoes de inteiros neste paiz (Sunkery no Canard) he de fazer
quatro partes iguais. ...” Se este parentese e do original, nao se
explica que o autor, versado como era em marata, limitasse so a
Sunkery uma pratica que devia saber ser geral em todo o Maharastra
e mesmo fora ; e ele proprio diz adiante a pg. 45 que isso e “ costume
dos indianos”.
II 0 alfabeto que indica como usado no pais e o marata e nao
o canares em que naquela regiao escrevem o concani.
III Todas as formas gramaticais dadas sao de Salsete (Goa) e
nao se encontra nenhuma de tantas peculiares ao Canard e que sao
mencionadas por Noronha e Maffei (n."’ 18 e 19).
IV Os poucos nomes geogrdficos citados sao todos extra-canareses,
como Goa, Bardes, Anjuna, Bombay, Surrate.
V Se Fr. Francisco Xavier era Carmehta descal^o, nao podia ter
vivido e aprendido o concani em Goa, donde os Carmelitas tinham sido
expulsos em 1702 ; nem podia ter sido missiondrio em Sunkery, que
pertencia ao Padroado portugues ; nem podia escrever em portugues
para os sens colegas carmelitas, que eram estrangeiros.
Todavia podemos continuar a designar o livro como a “ gramatica
do missiondrio italiano ” em vista dos tais ‘‘ resaibos ” da lingua
italiana.
8 — “ Francisco Jose Vieira, Desembargador da Rela^ao (1809-
1818). Eeduziu a regras e preceitos gramaticais o idioma proprio de
Goa, segundo afirma o Sr Manoel Felecissimo Lousada de Araujo,
tambem Desembargador da mesma Relay.ao, na Segunda Memoria
Descritiva e Estatistica das Possessoes Portuguesas na Asia, publicadas
em 1842 nos Annais Maritimos e Coloniais pg. 451 ” (Rivara).
9 — “ D. Fr. Manoel de S. Galdino, Arcebispo de Goa (1812-
1831). Tendo-se aplicado ao estudo da lingua, chegou a pregar nela,
segundo dizem. Hd tradifao vaga de que tambem compuzera uma
gramdtica, mas ninguem a viu nem dd outra noticia dela. Talvez
possuisse algum exemplar da que agora reimprimimos (de Estevao)
ou copia de alguma outra e dai viesse a fama de ser obra sua ”. Rivara.
10 — Xotas Gramaticais dadas por Francisco Luis Gomes como
suplemento a Gramdtica de Estevam, na edi 9 ao de Rivara (1857).
materia na maior parte extraida do Ms. do n.° 7, sendo, contudo,
algumas originais e vahosas.
724
JI. SALDAXHA —
A estas dez gramaticas, indicadas por C. Rivara, podem-se agora
acrescentar mais as seguintes 17 :
11 — “ Arte e Doutrma Christa em lingua Canarim.” Ms anonimo,
existente na Sch. O.S. (Marsd. Coll.. 11.280). Esta tnincado, tendo
perdido as ties priineiras folhas. Nao tern data ; mas certas tabelas
que veem no fim levam a supor que tenha sido escrita em 1615 menos
19 anos = 1596, ou 23 anos antes da morte de Estevao. Em vista
de muitas analogias, parece ser baseada na Arte deste jesuita, se nao
e a propria Arte, ou copia, da mesma, feita antes de ser esta acrescentada
por outros padres, tanto mais que e acompanhada de um “ Confessionario
e Boutrina em Lingua bramana canarim ordenada de dialogo que
se ensina Pellos Padres da Companhia de Jesus na India Oriental ”
o qual em grande parte se conforma, no texto e ate no titulo, com o
catecismo de Estevao.
12 — ■ ' Arte do Canary m, composta pelos nossos Padres e tresladada
polla mao do clerigo Antonio da Silva, Bramane de Margao. Tenba-se
em muita estima porque este clerigo a tresladou com muita curiosidade
sem erros, pa. o Pe. Antonio de Magalhaes bindo a Roma por Proc.®*’
a levar como levou e a tornou a trazer o Pe. Fran.° Caru.° bindo la
por Procurador.” No British Museum (Add. Mss. n.° 10.359). Tambem
tern grande semelban§a com a de Estevam, embora trate primeiro das
declina 5 oes e so no fim, da fonetica.
0 mesmo Ms. contem ainda : a) Alguns confisionarios na lingua
da terra mto. bons ; b) A doutrina cbrista abreviada com todo essencial
que bu Xpao (christao) deve saber ; c) A Cartilba e doutrina toda
int.® tresladada da impressa composta pello Pe. Tbomas Estevao ”.
13 ' Jnnua Indica, sive Pro Concanica et Decanica Linguis
Manuale Indias Ingres.suris per opportunum P. Ignatius Arcamone e
Societate Jesu Indiae Candidatis Socivs Bedicat.’ * Ms., provavelmente
original, existente na Biblioteca Nacional de Lisboa, (Reserv.
n. 3.049). Nao tern data, mas deve ser do meado do seculo xvk-
em que Arcamone esteve na India. Trata primeiro do concani de
Goa e em seguida expoe os ponto.s em que este se diferen^a do marata
decanico.
' Emijor.i o seu nome, como italiano, f6si,e Arcamoni, na India asainava
Ardh)amone. Viyeu mais de 20 anos iin Salsete (Goa) e publicou em conoani— 1,
Exphravoc.s dos Evangelhos Dominieais do todo o ario. Saynlleu Varu.warke etc ;
e 2, um Comentario sobre o l^urgatorio. Eoi ile ipic em 16.16 deu o visto cm portuguOs
assinando Irjnatio Archamone as Jardim da.i failures, era Concani, do Miguel
Indica as estilo da epoca em que havia Janun Graeca, Janua
HISTOEIA DE GRAMATICA CONCANI
725
14 — “ Gramnmfica Linguae Canarinae, quam gentiles Goani et
circumjacentes Ethnici inter se loquuntur.’’ por Charles Przikril.
Foi escrita em Goa no meado do sec. xviii. Przikril, que era jesuita,
natural de Praga, esteve em Goa desde 1748 ate 1759. em que,
pela extin 9 ao da Companhia, foi mandado preso. juntamente com
outros colegas, para Lisboa, onde foi posto em liberdade a pedido da
rainha austrlaca Maria Teresa, regressando a sua terra.
15 — “ Prosodia della Lingua Canarina pelo Pe. Diogo do Amaral
S. J. (sec. xviii). Seria interessante. porque nao existe outro trabalho
sobre a prosodia e metrifieaqao em concani.
16 — ‘‘ Grammatica MarasHa a mais vulgar que se pratica nos
Reinos de Nizamaxa e Idal.xa ; oflferecida aos muitos (sic) Reverendos
Padres e Missionaries dos ditos Reinos. Em Roma mdcclxxviii.
Na Estamperia da Sagrada Congrega 9 ao de Propaganda Fide.’" 8 .°,
45 pags. Desta edi 9 ao existem dois exemplares em Londres : um na
S.O.S. e o outro no India Office. 0 livrinbo foi reimpresso com o mesmo
titulo em “ Lisboa na Impressao Regia ; Anno 1805 " 51 pags. Nao
da nenhum prefacio ou advertencia que lance alguma luz sobre o livro,
ou o seu autor ; so a ultima pagina traz o Decretum Sacrae Con-
gregationis Generalis de Propaganda Fide, de 26 de Janeiro, 1778,
mandando-o imprimir “ Europaeis cbaracteribus ’’ na Imprensa da
Congrega 9 ao, donde parece que o autor e jesuita. mas o li\To nao e
mencionado na bibliografia jesuitica de Sommervogel.
17 — “ Esquisse Grammaticale de la Languc de Goa, por Johannes
Gonsalves.” Nao tern a pagina-fachada, parecendo ser separata de
alguma revista. Nao se sabe o lugar nem o ano da publica 9 ao (sera
1879, como indicou o catalogo de uma li\Taria francesa i) Nao se sabe
tanibem se o autor e indiano, se europeu. Por algumas incorrec 96 es
que diz nao pode ser blho de Goa. Come 9 a por afirmar que “ I'idiome.
goense est compose de maratte, de Sanskrit et de persan ” (!).
18 — " Konkani Grammar, by Rev. J. Pio Noronha.” Mangalore.
Ms. de que existem duas copias no British Museum (Oriental Mss
2.730). Uma destas copias, pertenceu ao Dr. A. C. Burnell, como
se ve da seguinte nota do seu punbo. escrita na primeira pagina.
■■ This Konkani Grammar was written for me by a Catholic priest
at Mangalore, named Noronha ; he was a konkani by race and had
devoted much time to the study of his native language (1873—4).
A. B.” (Sold to the Museum 18 Jan., 1884.) F um trabalho extenso,
de 553 fols. mas, apesar de ser moderno. e escrito sem o deffido
criterio gramatical.
726
JI. SALDASHA —
19 — “A Konkani Grammar, by Angelus Francis Xavier Maffei,
S.J. Mangalore, Basel Mission Press ; 1882.'’ Esta gramatica, alias
bem coordenada, nao podia deixar de conter certas deficiencias,
escrita como foi depois de um estudo so de ano e lueio da lingua,
deficiencias que foram reparadas pela obra seguinte ;
20 — “A Sweet Voice from the Konkani Desert (Konkni Ranantlo
Sabhit Sundar Talo). by A. F. X. Alaffei S.J., Mangalore. 1884(1).'
E uma gramatica complementar e correccional da anterior.^
21 — Gomnntaki Bhashci. Elementos graniaticais dados eiumarata
por Ramacli. Bh. Gunjikar no seu Sarasvati-Mnndala, Bombairn. 1884.
22 — Elements of Konkani Grammar, by Sten Konow. no
Linguistic Survey of India, de Sir G. Grierson. Vol. Marathi. 1905.
23 — Konkani Bhdxechem Lagha Vydkarana, por G. 0. Pires.
Bombaim. 1907. E a primeira gramatica organizada eni concani
e para os naturais ; mas, tendo sido escrita em carateres balbodha
e sein reclame, e completamente desconliecida. 0 autor considera
esta sua tentativa a primeira para a compila 9 ao de uma gramatica
da lingua concani ” naturalmente porque nao teve ciencia de tantas
outras que o precederam. 0 autor. que muito se empenhava pela
cultura da sua lingua vernacula por parte dos seus patricios. publicou
tambem nos mesmos corateres, as prirneiras cartilhas para a leitura.
A sua gramatica, decalcada sdbre a Marathi Lnghu Vydkarana. de
Dadoba Panduranga, e, como esta, muito elementar.
24 — Concanni Bhaxechem Lagha Vedcrann. por Duarte Francis
Dantas. Savantvadi 1910. Em concani, carateres romanos segundo
o sistema da Uniao Goana (vide adiante). Introducao por Jerome
A. Saldanlia. E tambem elementar de 36 paginas.
25 — Elementos de Gramatica Concani. por Mons. S. Rodolfo
Dalgado. Xao chegou a ser impres.sa pela antecipafao da morte do
autor. que, em 1922. ofercceu o Ms. a Biblioteca Xacional de Xova
Goa. onde ao presente .se encontra. E a unica gramatica cjuc. alem
de ser correcta e cientificamcnte coordenada e suficientemente deseii-
volvida. segundo o piano da gramatica marata de Xavalkar, O autor
teuton no fim um estudo sobre a gramatica historica da lingua, que
nao poude continuar pela gravidade da .sua liltima doenya. Os
cxeniplos sao escrito.s em aifabeto balbodh e no dos orientalistas com
certas modifica^oes.
^ Maffei era tiio apreeiador da Jm^ua, que liie rhamou “ sweet voire " e na
"raniatioa e&creve : '* Konkani is a ri( h ami heaiitiful language, although at present in
an lirnuhle state, bt-cause it is far more perfect tlian many European laniruagcs ;
yet !t Is altogether uncultivated and appears to be the most imperfect."
HISTORIA DE GRAIIATIC'A CONCANI
727
26 — Elemenios Gramaticais de Concani. pelo Con. J. de Sousa,
Lisboa. 1929. Compreende so a fonetica e a morfolooia e alsiuns
apontanientos avulsos sobre a sintaxo.
27 — ” Grama lien da Limjaa Concani. por V. J. Janim Rangel;
Bastora. 19.33." L escrita cm portujTues e conoani ein earatere.s
romanos sewundo o sistema popular.
Eis a relarao que se poude orijanizar com o au.xilio de elemontos
colhidos em bibliotecas, E po.ssivel (pie tivesse havido inais wramatiea.s,
hoje de todo de.sconhecidas ; e me.smo nem todos os dados sao
■suficientes para a reconstitui^ao biblioizrafica. Assim. por exemplo,
um Codex da Bibl. Nac. de Li.sboa da como ao tempo oxistentes. eutre
os livros do Convento da Cruz dos Milapres de Goa. as se<;uintes
graniaticas : a) " De Alvares, duas Artes : h) Duas Artes da lingua
da terra, uma impre.ssa e outra ^Is. “ A informa(;-ao e tiio manea, que
nao se sabe quetu eram os autorcs das duas artes da lingua da tiura.
nem de que lingua eram as artes de M.’’' Alvares. >Seriam a do
n.° 5, que era em “dois Li\Tos'‘ ? Mas esta era eserita por Simao
Alvarez e seu pai Louren^o .Vlvarez e niio por E provavel que
se relira ao celebre jesuita Manuel Alvares. que deixou alguns trabalhos
sobre a gramatica latina.
Em facto a notar 6 que. apezar de tantas graniaticas e algumas
delas impressas. nenhum dos autores -com a excep^ao dos quatro
liltimos — cita qualquer dos seus predecessores ou mostra ter deles
conliecimento. produzindo portanto um trabalho independente. So
o missionario portugU(?s (n.’ 2) fala. as vezes. da “ Arte de Goa
parecendo referir-se a Arte de Estevao. sobre eujo piano 6 coordenada
a .sua.
Orgaxizacao gr.vmatic.\l : Os primeiros autores, nao tendo
nenhuma obra indigena que Ihes servls.se de guia. seguirani o sistema
de gramatica latina. tanto mais que e.sereviam para missionarios
europeus.
Dividiram. em geral. a .sua Arte em tres partes principals : ■' A
primeira parte — diz Estevao — trata do alfabeto e das nossas letras
(jue respondem aos carateri’s da terra e dos acentos e pronunciaijao.
.\ segunda das oito parties da orac^ao. A tercoira contun a syntaxis."
A maior parte ddles nao trata da composic^'ao e deriva9ao. de que o
primeiro a se ocupar foi talvez Arcamone.
Xa Fonttica. o primeiro problema em que deviam ter esbarrado e
o do alfabeto e ortograba. Xao po.ssuimos monumentos literarios
anteriores a domina^ao portuguesa. para .sabermos se a lingua teria.
728
II. SALDAXHA —
nessa epoca, manifesta^ao grafica e em que carateres, i. e., se maratas
ou canareses. Estevao fala das " nossas letras que respondem aos
carateres da terra ” e especifica ; ” nestas terras os carateres nao sao
letras senao silabas ' ' o que se aplica tanto ao alfabeto marata como
ao canares. que devia ter sido familiar ein Goa na dominajao de
Vijavanagar, como o provam alguns documentos concanis dessa epoca,
escritos em carateres canareses, que se encontraiii em arquivos
comunais. 0 Missionario italiano, porem. da expressamente o alfabeto
marata como o vernaculo da terra. Mas qua! seria a lingua escrita i
Seria de facto o concani ou o marata ? Em nenhuma destas linguas
existe hoje qualquer produ<;’ao literaria da epoca pre-portuguesa.
escrita em Goa. E possivel que os antigos liindus de Goa — ao iiivez
do que fazem ao presente — nao .se dedigna.sem do sen vernaculo. pelo
menos para a escrita nao literaria. Mas, desde o seculo xvii. com o
progresso politico dos maratas e dos portugueses, os hindus se foram
encostando a cultura marata e os cristaos a portuguesa, ao ponto de
se envergonliarem. uns e outros, da sua lingua materna. que dcou
completamente abandonada. sem ensino e re.strita ao uso coloquial.
Ao sul de Goa. porem, tanto os naturais como os emigrados goeses.
achando-se longe da influencia marata. ainda escreviam o concani.
mas em carateres maratas, como o ate.sta o mis. ital. (se e que a sua
gramatica foi escrita no Canara) e urn atestado passado por tros
brumanes de Cochim, oriundos de Goa. abonando o Hortufi Indicux
MaJahiiricxs, do holandes H. Van Klicede. Aimsterdao. 1678, o qual
at(>stado e escrito em ' lingua bramana " cm carateres balhodh e vein
publicado no mesmo livro. Maffei. que oscreveu em Mangalore,^ em
1882. tambem afirma que ” antigamente o concani era cscrito em
hdlhodh e as vezes em modi ; mas hoje e geralmente u.sado o alfabeto
canares ”.
Desde os ultimos dois scculos. <a lingua esta sendo escrita quasi
exclusivamente pelos cri.staos e em tres carateres : entre os goeses,
os romanos, no norte os maratas e no sul os canare.ses.
Alem destes alfabetos, indicados pelos gramaticos, o concani teve
mais tres, que, por serein acualmcnte mais ou menos seguidos, merecem
ser aqui consignados ;
a) 0 alfabeto Marumo : No ultimo cjuartel do sec. xix, quando
alguAs filhos de Goa, espalhados pda India Britanica, recoiiheceram
a necessidade de proporcionar leitura concani a sens patricios ignorantes
de portugues e ingles, fundaram um pcriodico intitulado UdentecJiem
Snllok, (0 Lodam do Oriente), para o qual o .sou fundador, Eduardo
HISTORIA DE GRAMATICA COXCANI
729
de Sousa, convencionou uni sistema alfabetico. a que chamou alfabelo
Mariano em louvor da Virjiem Maria " ; o qual. einliora hoje
adoptado por uma revista, nao captou o apoio do publico por causa
de seus carateres especiais, inacces.siveis a tipografia vul^cir.
h) 0 alfabcto da Vnido Goana : Como a escrita popular paroccsse
pouco cientifica e as vezes arbitraria, a as.sociayao Unido Goana. de
Bombaim, tendo por objectivo zclar pelos iatere.sscs dos cmi^rados de
Goa. nomcou. em 1905. uma comis.sao pora apresentar um novo
projecto. que csta indicou e o qual, iiao exi^iiido tipo especial, foi
aprovado e e ainda seijuido por al^in.s publicistas ; mas nao logrou
a TOlgarizagao. principalmente por ter mantitlo o a breve, nao
mudo. representando-o por a italico. ao estilo dos missiouarios. o que,
alein de dificultar a escrita e impressao, contraria a tendencia
popular, que nao admite es.sa vogal, identificando-a para todos o.s
efeitos com o.
c) 0 alfabeto popular : 0 povo. portanto, e especialmente os que
ignoram o marata, continuani de.sde o ultimo seculo a usar o que se
pode chamar o alfabeto popular, que nao e mais que o dos
ini.ssionarios e da Uniao Goana, sendo eliminado o a breve, quo e
sempre substituido por o. E este o adoptado pela gramatica n.“ 27.
Os primeiros missiouarios, destinando os .sens trabalhos aos colegas
europciis, escolheram naturalmente o alfabeto romano mediante
certas adapta^des. porque, como diz Arcanione ; " Verum e.xperientia
didici : quibus literis ab iucuntc aetatc instituimur por ea.sdem
facillime linguarum nomina verba coeteraque vocabula memoria
retineri." Foi a primcira aplica^ao deste alfabeto as linguas
indianas. que depois seria dc.sonvolvida o aperfoicoada pelos
orientalistas. Mas. alem das gramaticas e vocabularios para o uso de
europeus. os missiouarios compu.soram ainda livros do caratcr roligioso
para a edifica^ao dos neo-cristaos. os cjuais tambem cram escritos em
carateres romanos, tanto porque o sistema de onsino visava a ilirigi-los
para a cduca^ao luso-cri.sta. desviando-os da instru^ao vernacula,
ministrada polos hindus. e tida como prejudicial para esse tim. como
porque na iiistrucrm popular, proporcionada .so nos convcntos o
nas escolas paroquiais. o onsino de leitura, escrita e miisica. era todo
em portuguds. para o quo prccisavam do alfabeto romano. E assim. os
descondontos dos jirimoiros convertidos ja nao estudavam as lingifts
do pais, nem coiihcciam o alfabeto indigeiia. facto C[Uo se da ainda
hoje cm Goa. ondc (puisi se pode dizor quo nonhum cristao froqucnta
as escolas primarias maratas.
voL. \air. p.ii;ts 2 and ;i.
47
730
HISTOEIA DE GRAMATICA CONCANI
A adapta 9 ao do alfabeto romano deve datar portanto dos primeiros
tempos das conversoes. 0 proprio Estevao deixa entender que o
sistema que seguiu nao foi por ele inventado, mas apenas adoptado
o que ja existia, porquanto escreve ; “ Aqueles que nestas partes
estudam a lingua da terra (europeus) nao a escrevem comumente
senao na nossa letra ” com a ajuda de “ acentos e letras dobradas e
aspira 9 oes e com as regras que ao diante se verao ” . Foi este o sistema
seguido em Goa para todas as publica 56 es desde o seculo xix com rara
excep 5 ao.
Escolbido o alfabeto romano com base na promincia portuguesa,
ate onde a fonetica das duas linguas o permitia, supriram a sua
deficiencia com certas convengoes indispensaveis, que, embora nao
satisfizessem a todas as exigencias filologicas, nem por isso deixavam
de corresponder de modo geral as necessidades praticas da lingua,
representando todos os fonemas realmente em uso na fala popular,
ao ponto de o orientalista Burnell * entender que o sistema era “ muito
mais perfeito do que os sistemas defendidos 200 anos depois em
Calcuta ” (pelos ingleses).
A ordem alfabetica era a latina, sendo enxertadas no lugar com-
petente as respectivas aspiradas e cerebrais. Segundo Estevao, as
nossas letras que podem responder aos carateres da terra sao as
seguintes :
a, a, b, bh, c, ch, chh, d, dh, dd, ddh. e, gue, gh, h, i, y, k, 1 , Ih,
11 , m, mb, n, nb, nn, o, p, pb, q, qb, r, rb, s, sb, t, tb, tt, ttb, u, ub,
y, z, zb.”
A razao de y figurar duas vezes e a de ele representar 7 longo e a
semivogal ija
Para confronto dos diferentes sistemas damos a seguinte taboa
alfabetica :
Deste quadro se ve que todos os gramaticos, antigos e modernos,
tirado o ultimo (n.° 27), sao muito precisos em fazer distinqao entre
a e o vogais cuja promincia, mesmo em marata, se aproxima
e entre os cristaos de Goa quasi se confunde. A representagao, porem,
de « variou muito conforme o sistema. Estevao transliterou -o
por a italico, que, por ser mal feito, levou Rivara a sup 6 -lo alpha
grega. substituindo -o na sua edi^ao por 6. a imita 9 ao da gramatica
marata de F. Neri Fires, ne.ste a ital. que foram impressas todas
^ Bumel], op. cit.
QuADRO COMPARATIVO DOS PhINCIPAES AlFABETOS
i
!
t
th
d
dh
n
t
th
d
dh
n
P
ph, f
b
bh
m
y
r
V
s
s
h
1
•ou^Tj'ej\[
« P<«M.£2^ P ’ > X X ®,i3— I-
-•B^uatJO
'5- '5 ^ ~ ^ !Ic £ !>■. >1 — p- c- CO- m .rt j
•S9UOX* \\\
g. c- ^ :g g c Pa-a - 2 £ ^“ ^ - > '^ -g ” i
1
1
•'BUBO')
OBlUfl
•aB[ndoj
•soj'^no 9
1 JBdSB£) *J J
§*"^'^=3 = ^ = .^r ^ S x: ^
•niBA 9 isa
^ j-H^ 3 X X
•usSbx
WhOphs^pplw^PpirtflB-firfflyKlFlirts^lB-plKyfg
; cs" ic3 .13
i -cf ^
^o «o 'O ' ^ ^ ^
•s^'jsr^
-■B^uauo
C} e d 3
I !! li !i ii
' to ^
Qj ^ 2 ^ "5:
c5 i::5 •- i- 3 (3 M. It- <y O § 'S ^ S ^ ^ ‘ ^
•S9UOf • \\
<3 ’t,.
O
tC-Tr-r;
•VU'BOf)
oipTu;^
e c5
3 3
o
•jB|ndOjj
o
I
•sojjno 9 1
JBdSBf) 'J^ ,3'
•niBA9iS3
i w f p IP IP I? r pf iw
jh, zh I jh, zli
732
M. SALDANHA —
as obras de Estevao e doutros missionarios antigos. I Mis. port, porem
e Fr. Gaspar (n.'® 2 e 3 ) representaram-no nos sens Mss. o ” longo
com a carapu^a para baixo, a, e o breve com ela para cima. a ". Ouso
do acento circunflexo na India para a C[uantidade longa e portanto
anterior a Jones. 0 que e curioso e que a transliteragao reproduzia
fielmente, ate na prosa, a maniera do sistema de Jones e dos
orientalistas, todo a inerente, ainda niudo, e assim escreviam
bassata e bassata, embora se pronunciasse bassat e basta. Esta vogal
tinha jano sec. xvi o mesmo som que tern hoje, i.e. " tern a pronunciajao
meia entre a e o” (Estevam) e " melhor fora — diz o mis. port. — C[ue
se escrevesse ou ; vdchi = le tu e vneha ~ vai tii ; melhor fora qne
(este) se escrevesse voucha ” .
Actualmente os goeses cristaos, nao costumando estudar marata,
perderam por completo a distin9ao entre a e o, c|ue sao pronunciados
da mesma forma = o, aberto ou fechado ; donde o sistema orto-
grafioo popular representa-os ambo.s sempre por o.
Uni outro ponto em que todos esses gramaticos antigos eram muito
cuidadosos era na representa^ao das eonsoantes cerebrals e das
aspiradas, porque “ trocando uma so letra ou acento ou aspirayao, nao
somente se corrompe a escritura e pronunciacao, mas tambeni se muda
as vezes o sentido ; e aqueles ejue nos ouvem nao nos entendem."
(Estev.) Para as cerebrals recorreram a duplicafao dos carateres e
para a aspira^iao ao A, processo que b ainda lioje seguido em Goa,
como mais pratico e nao exigir tipo especial ; Arcamone porem no
seu Ms. representa as primeiras cortando a letra por uma linha, per
transversani linvam inciaac.
Os dois sons de ^ (j e z) sao transliterados so por r : (uzo e ti(ze»i
= tujem ; e o mesmo fizeram com re.speito ao cujos dois sons (cA e
tch) representaram por ch. c, f, j, e A teoricamente nao figuravam
ne.sse alfabeto. “ Porque esta letra r ou a pronunciamos como ca. 90,
9i, 90, 9U. E entao sempre nos servimos de s. ou como ca co cu. E entao
deviamos usar de A, cuja aspira9ao e AA. de modo que c nao serve
senao quando pronunciamos como r/ia. die, chi, cho. dm. E por isso
se pde neste assy ch. Tambem /’ nunca serve, porque supre-se muito
bem CO pA, aspira9ao do p. Tambem q nao se mete meste alfabeto.
porque melhor se sati.sfaz com A. cuja aspira9ao e AA, como dito e.
Mas, contudo, como as na9oes dihcilmente recebem letras e modos
do escrever nao usado.s, deixando a parte esta letra /. quo nunca serve,
usamos a cada passo de c, porque melhor nos vein a mao que esta
letra A e quando vein sua aspira9ao, que devia .ser AA, usamos de qh.
HISTOKIA DE GRAMATICA CONCAXI
733
De modo que avendo de escrever ka, ke. ki. ko, kii, escrevemos ca. que,
qui, CO, CH. E pollas aspira56cs qha, qhe, qhi, qho. qhu. E o uso de kh
deixamos polo vocabulario do canaritn, que sem isso teria grande
confusao " (Gram. n.° 12). O mesmo diz Estevao. Mas na pratica
escreviam portaiito ca. que. qui, co. cu e qha qhe, qhi. qho qhu, E es
ta conven^ao geralmente seguida em Goa por corre.sponder ao alfabeto
portugues e foi por i.s,so que o sistema da Umdo Goami a adoptou por
conteinporiza^ao .
Embora o concani possua todas as tre.s sibilantes, os nossos
gramaticos .so as transliteraram por .s — o -= o Entre
vogais o s era reduplicado. ss. segundo a fonetica portugue.sa, vg. dssu.
Moefologia : n) Declinncdes — Os gramaticos mais antigos
admitiam, a maneira do latim, 6 casos com a mesma designacao e
ordem ; mas isso era mais para o estudo da siiitaxe do que para a
declinafao. para a qual davam so dois : o ca.so recto (nom. e acu.s.) v.g.
vatta ; e o obliquo vatte (gen. dat. voc. e abl.). que era o proprio tema.
Reconheciam que as declina§6es concanis nao cram iguais as latinas,
i.e.. que. ao contrario desta lingua, as terminagoes eram senipre as
mesmas para todos os nomes, variando so a forma tcmatica ; e por
isso. 0 ital. limita-se. nas suas declina96es. a dar apenas o nominative
e 0 tema ou caso obliquo. ao qual se juntariam sufixos ou posposi^des
para exprimir as circunstancias.
0 mesmo autor e Arcamoni identificam o genetivo com o caso
obliquo, antecedendo assim de alguns seculos a teoria moderna, que.
tambem. atribui a formafao do tema ao genetivo sanscritico. 0
primeiro gramatico a adoptar o n.° dos casos sanscriticos foi, para o
marata. Carey e, para o concani. Malfei. que foi seguido por todos
os autores posteriores, com a excep9ao do ultimo, que ainda seguiu o
sistema dos casos latinos.
Nao sendo a forma^ao tematica tao nitida como em latim. o n,° dos
temas. e. portanto. o das declina^oes, variou. como em marata, de.sde
tres ate nove. conforme os autores. Quern, porem, excedeu a todos
foi Xoronha (n.° 18). cuja gramatica. destituida de espirito sintetico,
da 5 declinagdes para os nomes prdprios e 8 para os comuns, que por
seu turno sao divididos em animados e inanimados ; cada declinafao
abrange 3 variedades. que dao ao todo 30 variedades de declinacoes (!)
que nao merecem a pena de serem aqui expo.stas. Arcamoni admitiu
3. segundo os 3 genero-s.
Mas. ja no sec. xvii. tanto Fr. Cristovam como Fr. Gaspar se
mostram mais cienti'ficos admitindo apenas os 4 temas primitives
734
M. SALDANHA
a, e, y = 7, it, entrando no primeiro os sens derivados ea, ia, e ua
e no segundo ye e ve. Esta teoria de 4 declina^oes era, portanto, ja
ensinada em Goa ha tres seculos. Maifei, Dalgado e o italiano
distinguem entre ea e ya, por causa da diferenga da pronuncia. Com
efeito, se em marata ghodydm e tema plural tanto de ghoda. como
de gho^, em concani a pronuncia nos dois casos e bem diferente.
Pronom-es : Sao di\'ididos por todos os antigos ate o italiano, em
primitivos, como hanv, fit, to, ho, zo e derivados destes hanvem am’chem
etc., sem se advertirem de que estes sao casos obliques daqueles.
0 ital. e o da gram, marastta admitem vocative para turn ; os outros,
h. maneira dos autores sanscriticos e pracriticos, nao o admitem ;
Arcamoni diz terminantemente “ vocative caret ” .
b) Verbos : Para o ital. “ a doutrina dos verbos parece um
labirinto ” . 0 verbo no que toca as vozes — diz Est. — e active (trans.),
passive e neutro (intrans.). Reconheciam, porem, que nesta lingua,
como na maioria das sanscriticas, nao ha propriamente conjugagao
passiva ; mas que nos verbos transitives os tempos do perfeito “ con-
cordam com o paciente em genero e niimero, a maneira dos passives do
latim ” ou, como diz Arcamoni : “ . . . non sunt activa sed Passiva ; in
ceteris temporibus activa tantum. “
Contudo, dominados pela gramatica latina, interpretaram como
voz passiva a conjugagao intransitiva de certos verbos, cuja significagao
em portugues e reflexa ou mesmo passiva, v.g. act. bhagassitd = perdoa,
e pass, hhagassatd perdoa-se ou e perdoado. So um autor moderno
ensina que “ quasi todos os verbos transitives se prestam a serem
apassivados ’’ do seguinte mode ; act. hanv (tacd) lott’tam eu empurro ;
pass, to maca lott’td, sou empurrado, teoria completamente anti-
-gramatical.
Admitiam quasi todos uma so conjugagao por causa da uniformidade
das terminagoes, embora variasse a vogal tematica. Os modos eram
muitos : indicat., imperat., optat., subjunt. e infin. Chamavam
participio em rus, correspondente ao latino laudaturus, ao partic. do
fut. kortolo ; e os geriindios eram tres: o 1° em di (=laudandi)
soduncho ; o 2° em do { = laudando) soduno ; e o 3° em dum
(=laudandum) sodunco. Derivavam os tempos de dois temas : o
do presente soditam e o do perf. sodilo. Como nao existe em marata
e cm concani o presente do infinite, a designagao verbal costuma-se
fager em marata pelo subst. verbal em nem v.g. marnem, e em cone,
pelo supine em unk v.g. marunk. Os antigos gramaticos e
vocabularistas, porem, “ nao o nomeavam ao uso port, pelo infinite ;
HISTORIA DE GEAMATICA CONCANI
735
mas ao uso dos dicionarios latinos pela 1® pes. sing, do pres, indie."
v.g. martam.
Syntaxe : Alguns autores acharam a sintaxe concano-niarata
muito parecida com a latina ; “ SjTitaxis Concanica fere non differt a
latina," diz Arcamoni ; e o n.° 16 escreve ; “ Como esta gramatica
seja feita para os que sabem latim, excuso de expor as regras da
Sintaxe ; porque este idioma segue as mesmas regras da Gramatica
Latina : somente farei men^ao de algumas cousas remarcaveis."
A todas as gramaticas, com a exclusao das muito elementares, a
sintaxe tern merecido um estudo especial e niinucioso. Estevao dedica-
-Ihe 168 regras e o mission, portug. outro tanto. Quem, porem, foi o
mais copioso foi Fr. Gaspar de S. Miguel, cuja Syntaxis Copiosissima
da 205 regras, expostas eni 78 pags. Na maior parte as regras sao
acertadas e bem observadas. Os exemplos sao todos redigidos expressa-
mente para o caso e nao extraidos da literatura, que ao tempo era
parca e mesmo essa de origem estrangeira.
As restantes partes da ora 9 ao (palavras indeclinaveis) tinbam sido
bem estudadas e minuciosamente descritas ; as preposigaes eram com
razao chamadas posposigdes.
Tal e a bistoria de gramatica concani desde o sec. xvi ate ao
presente. Enquanto abundam tantas gramaticas descritivas, nao
apareceu ainda nenbuma que fosse bistorico-comparativa, baseada
em filologia moderna.
Ein parthischer Titel im Sogdischen
Von Hans Heinrich Schaedee
T N der sogdisrlien Version des Vessantanijataka (VJ) ^ erseheint vor
dem Namen des Helden. des Konigsohns SirS"s)i :
oder an seiner Stclle das Wort u-yspySr'k (cinige iMale inj^pyhr'y,
196, 222, 232, 349, 364; Vokativ u-yspi/Sr"' 52', iryspy^r' 1198).
Gauthiot sah darin einen zweiten Nainen de.s Helden, *Wi><pi8nrdk.
den er als sogdische Vinbildung von skr. Tiintulara (pali Yr.^»i»tar/i)
ansprach. Erst Benveniste ® hat festgestellt. dass es kein Name,
sondern ein Titel ist, dessen Sinn sich unzweideutig aus Z, 36 des von
F. W. K. Miiller ^ herausgegebenen Fragments eines sogdischen
Padmacintamanidliaranisutra ergibt ; ZK yiii'ir Zl iryspSri/t 't
rnyh^t' der Konig und die Prinzen und die ^Minister Er iiber-
setzte das Wort nunmehr mit ,, prince ” * und erkliirte es als
Komposition aus n'yxpin) .. tout “ und S'r- .. tenir Die Bedeutung
ist richtig be.stimmt. die Etymologic scheitert .schon daran, dass nur
-Sr- geschriebcn wird. Vielmehr liogt hier einer der Fulle von pseudo-
hi.storischer (.. inverser ") Sclireibiing Sr start s vor : v'ij.^p{ij)hr-
ist vivpss < *iispus ; und darin selie ich die .sogdisclie Wiedergabe
des bckannten partlii.schen, auch ins Mittelpersische iibergegangenen
Titels vispulir .. Prinz — Benvenistes Beobachtung hat auch iiber-
lieferung.sgeschichtliche Kon.sequenzen. Indem sie den Namen
Visvantara aus dem Text ausscheidet. liisst sie zugleich <las wesent-
lichste iiussere lloment der Ubcrein.stimmung der sogdischen Version
mit dem Ve.s.santarajataka der Pali-Sammlung und der Fassung im
Kandschur verschwinden. Dafiir nahert der Text sich tiusserlich mehr
der Geschichte vom Prinzen Siu-ta-na im chinesischen Tripitaka.^
Aber er ist keincsfalls so mechanisch aus dem (’hinesischen iibersetzt
wie manche der mit ilim zusammen in Tun-huang geborgeneu
^ Ed. R. Gautliint. JA. 1.
“ Sogdiyiert au.s skr. Smlniia (chin. SiH-tn-na).
^ Gramm, .ioijil., ii. 107 A. 1.
« SBPrA. Rt20, 4.
^ Im Glossar 236b ,, prince heriticr
Ed. Chavanncs, Cifiq cent ft apohujiu^ t xttnit^ tlu T r. rhin.. iii (1901), 362 ff.
738
H. H. SCHAEDER
buddhistischen Lehrschriften, die streckenweise erst durch Zuriick-
gehen auf die cbinesiscbe Yorlage verstandlich werdend Dass der
Ubersetzer Sanskrit verstand, gebt aus der Behandlung der indiscben
Lehnworte und Namen hervor.
Das sogdiscbe Wort, das sich hinter der Schreibung wysp{y)8r-
verbirgt, erscbeint in etwas anderer Gestalt aucb in einem der in
Bulayiq bei Turfan gefundenen Fragmente eines neutestamentlicben
Perikopenbuches.2 In Gal. 4, 1 steht fiir syr. jartd (gr. KX-qpovofiog)
sogd. wysjnvsy : visjncH, was der Herausgeber F. W. K. MiiUer mit
„ Haussobn “ iibersetzte ; er hatte also das Wort scbon ricbtig mit
aw. visa pudra verbunden, hielt es aber offensichtlich fiir eine blosse
Umschreibung der Bedeutung ,, Erbe “.® Nun steht im Text zwei
Zeilen vorher (Gal. 3, 29) fiir ,, Erben “ (Blur.), syr. jar^, ein
verstiimmeltes Wort, das sich aber nach Mt. 21, 38 (ST i, 21, 8)
eindeutig erganzen lasst zu ptr]q'n xw'rt, d. i. wortlich ,, die Verzehrer
des vaterlichen (Erbes) Der Ubersetzer wird nicht in einem Atem
fiir denselben Begriff seiner Yorlage zwei so verschiedene Ausdriicke
verwendet haben, wenn er nicht einen Wechsel in der Bedeutung
beabsichtigte. Dass er dies tat, zeigt der folgende Yers 4, 2. Im
syrischen (wie im griechischen) Text entspricht dem ,, Erben “ in
V. 1 der ,, Yater “ in v. 2, im sogdischen dagegen dem ivyspwsy der
xs’um. Dies Wort bedeutet nicht „ Yater “, wie MiiUer mit
Fragezeichen iibersetzte. Wie Salemann sah, ist es mit dem haufigen
xsywn- ,, Herr “ zu vergleichen und bedeutet gleich ihm an dieser
Stelle ,, Kerr ,, K6nig.“ *
Allerdings wird in den christlichen Te.xten sonst nur xsytvn-
(var. xsun-) in der Bedeutung ,, Konig “ verwendet, wie in den
^ Vgl. die neuerlichen Untcrsuchungen von Benvenistc und Demieville uber das
Buddhadhyanasamadhisagarasutra, JA. 1933, ii, 193-248, sowie von F. Weller
iiber das Dirghanakhasutra, Asia Major^ 10 (1935), 221-8, und das Vimala-
kirtinirdesasutra, ibid., 314-364.
2 Zum Litcrarischen vgl. F. C. Burkitt, The Religion of the Manirkees, 119 ff.
^ SBPrA. 1907, 264 ; der Text in syrischer 8chrift wiederholt Soghdische Texte
{= BT) i (1913), 82, 21. Danach 8alemann, Ize\ Imp. Akad. Xauk^ 1907, 540 s.v.,
der das awestische Wort anfuhrt, al^er das sogdisehc mit ,, Erbe “ ubersetzt. Ben
Auslaut auf s vergleicht Bartholomae fragweise mit der altpersisehen Vertretung
^ von air. IF. 22, 105. Zusammenstellung mit vispuhr ohne Erklarung des Ver-
haltnisses der Worte zueinander bei Jackson, IF. 25, 181 ; Junker, Frahang i Pahlavik,
103b (wo ,, chr.-sgd. vispusi'''' statt ,, MPT, vispu^i^^ zu lesen) ; Reiehelt, Gesch.
der idg. Sprachwissenschaft, ii, 4, 2 S. 48.
* Salemann, loc. cit., 542. Ihm hat sich Muller stillschweigend angeschlossen,
wenn er spater zu ST. i, 108, anmerkte : ,, Von ixsdvan oder xsevan wohl abzuleiten
der Titel, nicht Name, der Hephthaliten.“
EIN PARTHISCHER XITEL IM SOGDISCHEX
739
buddhistischen Texten ys'ywn-,^ wahrend xs’icji sonst ,, Macht “
beisst.2 Aber ausser an unserer Stelle lassen sich noch zwei sichere
Belege fiir xswn „ Kdnig “ nacbweisen, der eine in einem christlichen
Text, der andere in der Inschrift von Karabalgasun. In Job. 5, 27
(ST i, 63, 8) heisst es : xstvn qt'rt wyny qt ivn’t. wortlich ,, er
macbte ihn zum Herrn, dass er tue “ (sjt. aslteh dnehwe ‘abed, aber
gr. i^ovaiav ebojKev avrw [/cpicrw'] TTOLeiv). Sowohl das Yerbum
(air. Tear-) wie das Pronomen wyny, das ein akkusativisches, nicht ein
dativisches Objekt — das ware qw wyny s' — bezeichnet, schliessen
bier die Bedeutung ,, Macbt " aus.* In den von 0. Hansen ■* entzifferten
Brucbstucken der insebrift von Karabalgasun sind Worte fiir
,, Herrseber ,, Herrsebaft u. a. ungeniein biiufig. Neben 'ysyivn-
,, Herrseber “ steben 'ysicn-yh {'ystv'n-yh) ,, Herrsebaft 'ys wn-
d'r ,, Herrseber “ sowie die Adj. yswn-'k (Frgm. 1-6, Z. 13, yr^
'ystim’k 'rkh ,, \dele Herrsebertaten “) und 'ysw'ncyk-. Dazu komnit
nun in Frgm. 9 Z. 9 das Satzstiick ZI ms J/Y ctp'r kyr'nw 'ysw'nty —
was nichts anderes beissen kann als ,, darauf die Herrseber aus den
vier Kiebtungen (tiirk. tori buhm) . . Hier beisst also 'ys w'n-
,, Herrseber ’b Ausserdem aber wird xsavan ,. Herr. Konig vollig
gesicbert durcb das gleiebbedeutende p'txfwn Luk. 19, 17, 27 (ST i,
47, 11 ; 49, 11) : beide verbalten sich offensichtlieh zueinander wie
np. sah zu pdSiMh.^ Aber ich gehe darauf hier nicht weiter ein und
bemerke nur, dass xsavan ,,Herr, Konig “ die von St. Konow ® zur
Erklarung des Titels saonano sao auf den Munzen der Kusan und des
sakischen Herrscbertitels ssau (?«;<) postulierte Bildung von xsd(y)-
,, herrschen " mit -van bestatigt und auch nocb in dem nocb fiir das
6. Jh. in Ostiran bezeugten Titel sdra (arab. sdba, arm. saug, chin.
sao-ivu) erhalten ist.
\Vir kebren mm zu der Stelle des Galaterbriefs zuriick. Nacbdem
fur xsdvan in v. 2 die Bedeutung .. Konig “ festgestellt ist, ergibt
sich fiir vispnsi in v. 1 ,. Prinz ‘b Der sogdische Ubersetzer hat auf
eigene Hand die ganze paulinisebe Metapher aus dem Privaten ins
^ Daneben steht das ebenso geschriebene "'ysyxvn ,, Klage
- Belege bei Mulier-Lentz. SBPrA. 1934, 583c, s.v.
® Vgl. damit SBPrA. 1934, 533, 28 : yy $y xPwn t'br pr 'yc syqn ,, und er gab ihm
Macht im ganzen Palast.*'
• Joum. Soc. Fiovgr.y 44, 1930.
® Vgl. Bartholomae, Miraih. Mundarten, 3, 51 mit A. 1.
® ZDMG., 68 (1914), 93f. Ihm stimmte A. von Stael-Holstein zu {JRAS., 1914,
84 A. 2), unter Hinweis auf das ihm von Salemann gezeigte ar.sVzraTi, dem er aber noch
nach Muller die Bedeutung „ Macht *' gab. Vgl. weiter St. Konow, Ostas. Zeitsekr. 8
(1919-20), 220 ff., und Junker, Airestaolphabet 103.
740
H. H. SCHAEDER
Hofische iibersetzt und darum auch fiir die apitrope ( < IttLtpottol)
des svrischen Textes sinngeniass ,, Befehlshaber ” {frm nd'rty) ein-
gesetzt. Der gauze Passus heisst bei ihm : ,, Aber ich sage, solange
der Prinz ein Kind ist, ist er nicht unter[scbieden] von den Knechten,
obwohl er Herr ist ihrer aller ; soiidern er ist unter {sub) den
Befehlshabern und Hausverwaltern bis zu der Zeit, die der Konig ibm
gegeben (=gesetzt) hat."
Yon den drei Formen wyspySr-. ivyspSr- und ivyspws- haben die
beiden ersteii besser den Konsonantismus, die dritte besser den
Vokalismus der vorauszusetzenden Ausgangsform *i'ispus bewahrt.
In der dritten hat anseheinend ein Ausgleich des Auslauts der ersten
Silbe mit dem der zweiten stattgefunden, in den beiden andern eine
Umfarbung des Yokals der zweiten Silbe. Diese Veriinderungen deuten
darauf hin, dass das Wort deni Sogdischen nicht von Haus aus eigen,
sondern entlehnt ist. Theoretisch konnte man *i'ispus fiir die normale
sogdische Fortsetzung eines altiranischen *i'ispudm erklaren, das ira
Parthischen durch vispuhr vertreten wird. Aber dagegen spricht
zweierlei : einmal die Tatsache, dass iin Sogdischen, soweit es uns
bekannt ist, puBra ganzlich durch zdt- verdrangt ist,i sodanii durch
die nunmehr fest definierte Bedeutung ,, Prinz Denn diese ist
natiirlicli nicht genieiniranisch, sondern unter ganz bestimmten
geschichtlichen ^'oraussetzungen an einer bestimmten Stelle entwickelt.
Einen festen Punkt bedeutet die Form vispuhr, die kliirlich parthisch
ist ; sie ist ins iMittelpersische entlehnt worden (die echt mittelpersische
Form wiire *vihpus oder *vispiis). Sie ist jetzt in persischen und
parthischen manichaischen Te.xten reichlich belegt. So ist in einem
alten persischen Jlissionsbericht von einem zur Lehre Manis bekehrten
Abkdmmling des parthischen Hcrrscherhauses, dem Prinzen {vis}>uhr)
ArdajSan die Bede.- In einem jungeren parthischen Hvmnus wird
der Lrmensch vispuhr suhrdurzdhay Prinz und Herrscliersohn "
genannt.^ cbenso in einem spiiten persischen Hnnnus.^ Unter den
persischen Hymnentexten hndet sich ein Stiick, das an den tiirkischen
^ So winl ill fluni inaiiK iK'rsLsch-soirdisi-huii Fnitfiuent M 172 (Muller,
Hafi‘(hrhnftf/ir* >tf, n, UK”) f.) pprs. -pu'tnr durch so,cr«l. prztu'f wictlor^egeben ;
vi^l. fcrnor da.-i \fln’TU‘inandcr voii hyirtk „To< liter '' mid ztk ) ,, Solin ", VJ.
20, 31, 3ol f., 1140, 1176, 1485. In den alten Hriefen erscheint nur einmal die
ideoifr.iphiM he S( lireibiuiH Illl
- Andrea.s-Hoiminiz, SliTrA. I0.‘jr3, 303 mit A. 3 (tUizu Iranica 72).
® Andreas-KcnninH, 1034,877,5.
« Henning, AV,Y_;ir. 1033, 313. 318 Str. 10b. (Andreas — bei R(‘itzenstein, Das
mand. Buch vom Herm der Grosse 46 — ubersetztc ri-^puhr iinriehtig mit ,, Sohn
des Geschleehts
EIX PARTHISCHEE TITEL I>I SOGDISCHEX
741
Prinzen Ogiirtniis gerichtet ist uiid ihn al-s vispuhr nazuy sahrijaran
Jrazend, ,, zarter Prinz und Herrscherkind ” anredetd Eiii weiteres
Zeugnis, das deswegen wertvoll ist, well es sicli datieren liisst, ent-
halt der 7C2 n.Chr. gescliriebene Kolophoii des maiiicliaischeii
Hymnenbiiclies ; darin bndet sich eiii Geljet fiir den derzeitigen
Uigurenherrscher ,, nebst seiner Nacbkommenschaft und seinom
Herrsehergeschlecht, den Prinzen nnd Prinzessinnen n^dy tdxm
uS sakrydre i xves vispuJirdn visduxtdn.^ Besonders Interesse verdient
die bier m. W. zum erstennial })elegte Form risduxt .. Prinzcssin ",
die Bailey ^ in der wobl aus nietrischen Griinden umgebildeten Gestalt
risdnxtdn — im 4’Is u Ramin des Faxrl Gurganl (11. Jh.) belegt
hat. Sonst sind mir beide Worte im Nenpersi.schen nicht bekannt ;
nach F. Wolffs Glossar kommen beide im 8chahname nicht vor.
Wenn hiermit die parthische Bildung rispuJir — deren Aussprache
mit i an alien angefiihrteii Stellen (lurch die Schreibung iry.^pichr
feststeht — hinlanghch gesichert ist, so bedeutet das nicht, class sie
eine Neuschdpfung der Partherzeit sein miisste, sondern nur, class
sie dainals stereotyp wnirde. Ehe wir aber dieser Frage weiter
nachgehen. muss zuniichst der Ubergang von I'ixpuhr zti sogd. vinpni
niilier erkliirt werden. Dass wirklich Entlehnung stattgefunden hat,
■nird. wie schon bemerkt, durch das Fehlen einer Fortsetzung von air.
pudnt im Sogdischen und durch die genaue Ubereinstimmung der
beiden Worter in ihrer speziellen Bedeutung .. Prinz ” nahegelegt.
Was aber das Yerhiiltnis von parth. hr und sgd. i- anlangt, so gewinnt
man bier einen neuen Beleg fiir den zuerst von Junker ® an dem
A'erhaltnis von parth. {a)hriman und chr.-sgd. iinnw aufgewiesenen
^ Andreas-Hrnnins. SBPrA. 1033, 332. 2.'>.
“ F, W. K. MuUt'r, DupjieUihitt 0. Vorhrr scclmn BBPtA. 10iJ4, 330. Siilemann,
A7 /g/.. 1 . 33, Herzfeld, Aichao], Mitt. au^. Inin- {AMI)^ 7 (1034) 18, A. 2. gibt vcr-
schentluh <Ir- TA’-'iing ca'^puktdn id'i'luxtdn ; fine Stelle de^M‘lhfll Texte^, direr ibid.
33 A. 1. be'«prifht. veristclic uh ainlers : ddn ku antlfu tltn mdhinn prtb n^zOn
Dojiprtlil. 10, 107, hfls^t nifht ,. MMla>s cs Avunle uiiler den RfhLn<n!S'=:rhrifien ver-
niehirnd ", c-uiidern :?oda>s es. in der (.leineinde ein Kapital aut ZiUM-n \Mirde
^ liSOS. 0. 033.
^ i)a-' M ht>u in dfii alten llriefen belegte Bypwr (Npater ^yp' trr).dd'- tlen rhincsi>ohen
Kai.'Cr bczeii linet nnd >finem einhfimi*^! hen Xitel Vifu-t'^p entsprieht, erweist sich
Sfh(«n (Inrch das r statt •' als nieht eiht Vizb einstni'ilen Benvfni^te bei
Mlaktr. WZKM. 42 (1033) 262 f.
■’ Fimari-'i lu“ Jahrbuelier 3 (1023) 40 11. -sunne aiuh hinter der biiddh.-
sed. Schreibung 'thrnunr steckt (ander.s Kenveiii^tc MSL. 23, 1027. 130 f.), wurde
durth <lie bn Unehelt, Sorjh'I. lInhiUclDiJtmn'it* ^ i, 7 A. 3. 8 A. 2, angegebene
(deidiung die.>e-' \Vorte> nut dem ehin. Aquivalent vou ''kr. mnrn nnmittelbar
deutli<h und i-'t ziier-^t von Lentz, SBPrA. 1034, 377b, ausgesprochen worden.
742
H. H. SCHAEDER —
Ubergang Jir zu s, der zugleich den Charakter der durch hr bezeich-
neten Lautung wenigstens in einem Teil des partbischen Spracbgebiets
erbellt. Dass dieser Ubergang selber schon im Partbischen so gut wie
abgeschlossen war, babe icb unlangst ^ an dem in manicbaiscben
Texten aufgetaucbten Gottesnamen Sromv gezeigt, der aus *Srdsahrav
entwickelt ist ; das war nur moglich, wenn hr und s einander nabe
genug standen, um nebeneinander tretend Silbenelbpse zu ermoglicben.
Damit kann die Entlehnung von visjnis aus vispuhr als erwiesen
gelten.
Was nun die Herkunft und Bildung von parth. vispuhr angebt.
so biingt daran eine Eeihe von scbwierigen sprach- und kultur-
geschichtlicben Fragen, die bier nur kurz bebandelt werden konnen.^
Wie verhalt sich vispuhr zunacbst zu dem awestiscben viso pudra ?
Aw. vts (recte : vis) ,, Clan “ ist die zweite Stufe in der viergUedrigen
Reibe der sozialen Verbande, nach dem Haus vor dem ,, Gau “
und dem ,, Land “ ; sein Oberhaupt ist der vispaiti (ai. vispdti),
der ,, Herr des Clans Man hat angenommen, der viso pudra sei
der Sohn des vispaiti ® — aber das ist meines Erachtens schon sprach-
lich schwierig und steht ausserdem im Widersprucb zu den awestiscben
Aussagen. Der mj-thische Held ©raetaona, der Sohn des A^vya, ist
der einzige, der im Awesta als vuso pudra bezeichnet wird. Aber diese
seine Bezeicbnung tritt nicht absolut, sondern mit naberen Bestim-
mungen verbunden auf : der Held beisst ,, Sohn des adwyanischen
Clans, des heldischen Clans “ Yt. 5, 33 ; 17, 35, sein Vater betet,
,, dass ibm als Sohn des heldischen Clans ©raetaona geboren werde ’■
Y. 9, 7. Diese drei eng miteinander zusammenhangenden Stellen
erklaren also Bildung und Sinn des absolut gebrauchten viso pudra
nicht. Dies tritt nur an einer einzigen, aber den Sacbverhalt klarenden
Stelle auf, in Yend. 7, 43. Hier ist von den Arztgebiibren die Rede,
die nach der sozialen Stellung des Patienten gestaffelt werden. Auf-
gezahlt werden Priester, Hausherr, Clanherr {vispaiti), Gauherr,
Landesherr und die Frauen der vier letzteren, dann folgt der viso
1 I'ngarische Jahrburher, I .7 (193.5), Heft 4/5 (im Druck).
2 Die wichti^sten noueren Boitrage sind : Xoldeke, Gesrh. der Perser und Amber
(1879), 71, 437, 501. Olshausen, MonatJiher. Preuss. Akad., 1880, 354 ff. Darmesteter,
Etudes iraniennes, ii (1883), 1.39 ff. Hubschmann, ZUMG. 46 (1802), 326 f. Arm.
Gramm. (1896), 80 f.. 178. Christensen, L'empire des Sassanides (1907), 20, 23 ff.
Bartholomae, WZKM. 25 (1911), 251 ff. Xeuerdings besonders Herzfeld. PaikuJi GL
157, 170. ^.1/7. 1. 145, A. 1 ; 2, 20, A. 1, .32 ; 4, .54 ; 6, 74 ; 7, 18 ff. Bailey, BSGS.
6, 80, 953 ; 7, 72, 75 [Vgl. die Xachschrift],
* So Christensen, I.c. 20.
EIN PAETHISCHER XITEL IM SOGDISCHEN
743
pudra.^ Ware dieser der Sohn des vispaiti, so niiisste fiir ihn der Ansatz
einer geringeren oder hochstens der gleichen Gebiihr wie fiir den
insjmiti erwartet werden, Tatsachlich aber ist seine Gebiihr eine
hbbere und kommt der des Gauherrn gleich. Daraus geht unzweifel-
haft hervor, dass der rlsd pudra nichts mit dem vispaiti zu tun hat,
sondern sozial iiber ihm steht — mit andern Worten, dass vis als
Bestandteil beider Bezeichnungen nicht das gleiche bedeutet : vispaiti
ist das Oberhaupt einer vis schlechthin, visa puOra der Abkommling
der fls Kar i^ox-qv. Das fiihrt aber auf einen sozialen Zustand, in
dem ein einzelner Clan bezw. ein einzelnes Geschlecht unbedingten
Vorrang gewonnen hat, also auf die Monarchie. Erst seit der Durch-
setzung des Kbnigtums in Iran konnte die Bezeichnung risb pudra die
spezifiscbe Bedeutung gewinnen, die ihr an der Vendidad-Stelle
zukommt. Zugleich wird deutlich, dass bier der Ursprung des
parthischen vispuhr liegt.
Die soeben postulierte Sonderbedeutung von ns : Clan
Kar'e^oxpv, Konigsclan “ ist in den Achiimenideninschriften
reichlich bezeugt. Dareios I. spricht von seiner vid sowohl in dem
abstrakten Sinne des Kbnigsgeschlechts, seines ,, Hauses “ (so Bis. 1,
69, 71 ; NRa 53), wie in dem konkreteren der .. Residenz “ (so Bis.
[2, 16] ; 3. 26 ; 4, 66 ? Pers. c, Pers. e 24), aber nicht in bezug auf
ein bestimmtes Gebaude wie dida, hadis, apadana, tacara.^ In der
akkadischen Version steht dafiir gewohnlich bltu ,, Haus “. Der
,, Sohn des Hauses in diesem Sinne ist der achamenidische Prinz.
Dies erklart einen Passus zu Beginn des beriihmten Briefes der
jiidischen Gemeinde von Elephantine an Bagoas, den persischen
Statthalter von Jerusalem, vom Jahre 407, wo dem Empfanger Heil
und ,, Gnade vor dem Kbnig Dareios (II.) und den Sbhnen des Hauses
{bny byt') “, d.h. doch : den koniglichen Prinzen,® gewiinscht wird.
Aram, br byt' ist die Wiedergabe eines altpersischen *vida pu^a, nach
dem das parthische vispuhr gebildet ist.
Es ist danach wohl verstandlich. das br byt' unter die Pahlavi-
Ideogramme aufgenommen ist — aber bier erhebt sich nun eine
* Ua.s unsmnige i i s- /»(.'■ put der l’ahlavi-t)ln'rsetzung gibt nicht i fso puBnm wieder
(so Bartholomae, Air. ]yb., 1457), sondern eine Kontamination von rDso puBrdm mit
dem in der Hds. Jp 1 stattdesson uljerlieferten vJ-io li-^pfiitim.
2 kara hya iiddpatiy ist das in oder bei der viB stationierte Heer, die Garde —
keinesfalls ist viBdpatiy ein Adj. mit der Bedeutung ,, in Sippen organisiert,
hochadlig wie Herzfeld. AMI., 2, 32, will.
® So hat auch Ed. Meyer, Der Papyrusfund von Elephantine® 81, die Stelle
verstanden. Aram, br byt' hat nichts rait akk. mar bnni zu tun (gegen Herzfeld,
AMI. i, 145 A. 1).
746
n. H. SCHAEDER —
WO die vaspuJirakan regelmassig nach deni Konig, seiiieii Solinen und
seinen Briidern erscheinen.
Herzfeld hat neuerdings ^ die Form vispuhr fiir die arsakidische
und sassanidische Zeit iiberhaupt bestritten und will iiberall vdspuhr
lesen. Dass das nicht richtig ist, zeigen die vielen sicheren Belege fiir
vispuhr aus der manichaischen Uberlieferung, die vorhin angefiihrt
worden siud. Herzfeld stiitzt sich dabei auf einen nierkwiirdigen Text,^
der sich nach einer soeben erschienenen Bearbeitung von J. Tavadia ®
als eine Art Muster-Tischrede darstellt. Darin wird zwischen dem
Grosskdriig (sdhdtisah) und dem Grossvezir (vrtzrnk framdtdr) der
pus i vdspuhr i sdhdn farraxvtum ^ genannt : damit kann in der Tat
nur der Kronprinz gemeint sein. Aber was bedeutet der Ausdruck ?
Herzfeld iibersetzt in AMI. 2, 20 A. 1 : ..dem Vaspuhr-Sohn. dem
gliickseligsten der Konige,"' ibid. 7, 18 : ., Dem Infanten, Sohn des
(koniglichen) Hauses, dem glorreichsten der Konige." “ Keine der
beiden tibersetzungen ist haltbar. ,, Vaspuhr-Sohn " wiirde einen
Angehorigen des Hochadels. im Unterschied vom koniglichen Priiizen
bedeuten. Die M’iedergabe des einfachen pus durch .. Infant " ist so
unzulassig wie die Auflosung der Verbindung pus i vdspuhr ; die
Ubersetzung von vdspuhr durch ,, Sohn des (koniglichen) Hauses "
steht mit der fiir alle andern Stellen gesicherten Bedeutung des Wortes
im Widerspruch und ignoriert den Unterschied zwischen vispuhr
und vdspuhr. Mir .scheint, dass vdspuhr an dieser Stelle nicht als
Xitel, sondern als epitheton ornans zu verstehen ist ® : pus i vdspuhr
heisst ., der hochadlige Sohn"’, namlich des unmittelbar vorher
gonannten Grosskonigs.
Damit ist die parthisch-sassanidische Unterscheidung zwischen
vispuhr ,. Prinz “ und vdspuhr ,, Angehoriger des Hochadels " gesichert.
Herzfeld hat zweifellos Recht, wenn er als Lesung des Ideogramms
brbi/t in sassanidischen Texten viberall vdspuhr statt vispuhr
fordert.'^ Nur darf deswegen der selbstiindige Fortbestand von vispuhr
nicht in Abrede gestellt werden. Es handelt sich jetzt noch daruin,
» AMI. 1. ,32. 7. 19 A. 1.
^ Jamasp-Aaana, Pnhl. Tert.c, ii, 1.5.5 ff.
^ Eine Tiachrerle aus (Icr Zeit der .Sasanirlen, Verlag J. J. .\ugustin, Gluckstadt,
19:?.5.
’ Puhl T-xt^. 1.57, 2.
Tavadia, 1 c, 8, ungeiiau : ,, der Kronprinz. der gliieklichste unter den Prinzen,”
' So auoh, wie irh nachtraglich sah, Bailey, BSOS. 7, 72.
’ Danaeh ist auch bei Christensen, Empire, 20, 23, 28 f., 56 usw. zu verbessern.
EIN PARTHISCHER XITEL IM SOGDISCHEN
747
das spracliliche iind sachliche Verhaltnis dieser beiden Bejiriffe
zueinander aufzuklaren.
Nach der schoneii Beobachtung von Bartholomae ' pelit rda- in
vaspiihr auf *vdis-, Vrddhi von vis zuriick. Wenn also vispuhr der
8obn des Herren-, Fiirstenbauses, Prinz “ ist, so kann vdspuhr
urspriinglich nur bedeuten aiif den vispuhr beziiglich. vom vispuhr
abstammend Bartholomae bat hieraus die allein richtige. aber.
soweit ich seho, nnbeachtet gebliebcne Erkliirung des eigentlieben
Sinnes von vdspuhr{ak) gevonnen ; es bezeicbnet ,, die Mitglieder
des persiscben Hochadels. insofern sie Abkommlinge des Kdnigsbauses
ivaren oder daftir galten “ (loc. cit.. '254). Das bedarf nur einer geringen
llodifikation : die urspriingliche Bezeichnung der Abkommlinge der
koniglichen Prinzen wurde in spiiterer Zeit auf die ihnen gleichge-
stellten Angehorigen der hochsten Adelsfamilien iibertragen. Wann
das geschah, ist nicht bekannt ; in friilisassanidischer Zeit ist jedenfalls
der enveiterte Begriff im allgemeinen Gebrauch.
Daraus ist ein Schluss zu zieben. den ich bier nur kurz andeuten
kann. Der Hocbadel. dessen Angebbrige in parthischer und
sassanidischer Zeit vdspuhr biesscn. war schon in acbamenidischer Zeit
vorhanden ; er ging aber in erster Linie auf das sich rasch verzweigende
Konigshaus zuriick.^ nicht — wie vielfach angenommen wird — auf
die neben dem Konigsliaus stebenden Adelsgeschlecbter, soweit diese
sicli nicht mit jenem verschwagerten. In seinem trefflichen Abriss
der altiraniscben Kulturgeschicbte ® nennt A. Christensen ini
Anschluss an Andreas die Hiiupter der sechs grossen Familien in
acbamenidischer Zeit i-isputi's (die er iibrigens mit Recht von den
sechs Jlitverschworenen des Dareios unterscheidet). Aber diese
Bezeichnung ist nirgendwo belegt, und aw. vlspaiti bedeutet, wie
oben gezeigt, den Clanherrn schlechthin, nicht das Oberhaupt einer
Adelsfamilie.'* Noch weniger ist es angangig, die Tatsache. dass
Dareios von seinem Hause als von ., dieser I'iO *' spricht, dahin zu
1 ICZA'.U. 2.5 (inil). 2.-)l-4.
2 Man (lerike an die ^’e^^leh^unfr der Kadscharenprinzen im Persien des ll>.
Jh. Feth *Ali Schilh konntc ja in dem gleiihen Sinne ,,Vaterdes Vaterlandes genannt
■^erden wie August der Starke.
® In W. Ottos Hamihuck <l€r AUertum-^wi'^^en.-ichafty S. 256.
* Daran scheitert auch die Deutung voii arm. Hau&herr“( = ,, Oberhaupt
eines Fursten- oder Adelscreschlechts *') als „ Ubersetzung von rij^pati (Christensen,
Empire, 10 A. 3). Die in manichaischen Texten fur eine untergeordnete Gottheit
verwendete Bezeichnung visbeB {< vi<patt) sagt uber die Bedeutung des altiraniscben
Wortes nichts aus.
748
H. H. SCHAEDER
deuten, dass damit die Achameniden als einer unter melireren adligeii
Clans hingestellt werden soUend
Wie alt ist die Ableitung von vdspuhr aus vispuhr ? Man wird nicht
mit Bartholomae bis ins Uriraniscbe zuriickgehen diirfen. Denn sie
setzt die Auspragnng der Bedeutung ,, Konigshaus “ fiir vis (rid)
voraus, die wir zur Zeit des ersten Dareios vollzogen fanden.
Ausserdem aber musste ilir die Bildung eines festen Kompositums
aus vis und pudra bezw. deren Fortsetzungen vorangehen. Hier erbebt
sicb eine Scbwierigkeit. Das Awestische kennt nur die Genitiwer-
bindung inso pudra, kein *vispudra. Entweder muss man nun fiir
das Altpersische ein Kompositum *vidpuga (*vispuga) ,, Prinz “ imd
ein da von abgeleitetes *vdidpugi (*vdispuci) ,, Prinzensohn “
annehmen, die sicb auf mitteliraniscber Stufe in den partbiscben
Umbildungen vispuhr und vdspuhr fortsetzten. Oder man muss —
was bedenklicber ist — annehmen, dass derartige Vrddhibildungen
noch auf mitteliraniscber Stufe moglich waren und dass vdspuhr erst
gebildet wurde, als nacb dem Abfall der Casusendungen aus der alten
Genitiwerbindung *vida puqa in parthischer Umbildung ein
scbeinbares Kompositum vis-puhr bervorgegangen warA Aber diese
Frage ist von untergeordneter Bedeutung. Dass anderseits das
urspriinglicb fiir *vida puga geltende aramaiscbe Ideogramni brbyt'
von luspuhr auf vdspuhr iibertragen wurde, ist wohl verstandbch.
^ So Sarre-Herzfeld, Iranische FelsenreliefSy 16 .
^ Dafur dass noch im Parthischen die Genitiwerbindung zunachst erhalten
geblieben ware, darf man sich jedenfalls nicht auf die lu.W. zuerst von X. Marr
{Zap. vast. otd. 11, 166) aufgebrachte, bereits von Hiibachmann {ZDMG. 46, 327)
abgelehnte Kombination der armenischen Adelsbezeichnung sepuk mit deni un^
beschkftigenden Begriff berufen. Denn zwar hat Andreas (bci Lentz, ZII. 4, 300)
sie wiederaufgenoniraen und ,^^^epurh"^ (1. sepuh < *,H^purh) auf {ri)sepuhr — also
mit Obi. auf e im Vorderglied — zuruckfiihren wollen. Aber wo gabe es sonst eine
solche Anlautverstummelung iranischer I.*ehnworte im Armenischen ? Eine
hochst kiihne Deutung von sepnh hat Markwart in einer Annierkung zu seiner Schrift
uber den Ursprung des armenischen Alphaliets (1917) S. 10, A. 2, mitgeteilt, die ich
mit einigen Glossen wiedergebe : ,, Sepuh oder eigentlich ^sep'uh, wie das Adjektiv
spp'hakan [, besonder, eigen ‘] zeigt, ist eine Cbersetzung des iberischcn sep^e-ruli
[d.i. georg. sep'e , Konig " ctili , Sohn ']. Das zum Suffix gewordene -uh, fern.
-urhi, -uhi [vgl. Vaguhi , Konigin " nelren t'‘ag , Krone % Vagnvor . Konig *], entspricht
dem ap. puthra [d.i. p«fa] , Sohn ’’ Diese Erklarung hat nur eine Schwache : cm
alteres sep'uh. kann man aus sep'hakan nicht erschliessen, denn dies ist nur eine
Kontamination der Ireiden normalen Schreibungen sephahan und sfp'akan {Meillet.
Arm. Elementarbuch, 10). Ausserdem kann das Suffix -uh, -u{r)hi nicht auf ap. pu<;a,
sondern nur auf puhr, *puhrT eines rnitteliranischen Xordwestdialekts zuriickgefuhrt
werden ; wenn die Ableitung richtig ist, so wurde sie zeigen, dass der Schwund von
anlautendem idg. p im Armenischen nicht vor dem 5.-4. Jh. erfoigt ist. Die altere
Form *sepurh, aus der sepuk hervorging, ist mit Anlautswandel s > z ins Georgischc
entlehnt Worden ; sa-zepur-o „ eigen, ausens ahlt Deeters, Caucasica, 3 (1926), 82.
EIX PAETHISCHER XITEL IM SOGDISCHEN
749
Ich fasse zusammen. Mit der Entstehuag des Konigshauses der
Achameniden gewann in der hofischen Sprache das ererbte Wort fiir
,, Clan “ die Bedeutung des Konigshauses Sie ermoglichte die
Bildung ,, Sohn des Hanses “ ini Sinne von „ koniglicher Prinz
*vi9a pu^a, aram. hr byt\ reflektiert durch visa pu6ra Vend. 7, 43.
Dazu trat eine Ableitung ,, Prinzenabkommling zur Bezeichnung
zuniichst der Nachkommen der jiingeren Prinzen, dann auch des
ihnen gleichstehenden hbcbsten Adels. Seit der Arsakidenzeit leben
beide Worte in parthiscber Form, vispuhr ,, Prinz " und vaspuhr{ak)
,,Magnat“, nebeneinander. Aus dem Parthischen nnmittelbar oder
durch das Persische vermittelt ist vispuhr ins Sogdische iibergegangen,
wo es in buddhistischen und christlichen Texten in den abgewandelten
Formen wysp{y)8r- : vispaS- bzw. vispus- unter der unveranderten
Bedeutung ,, Prinz ” erscheint.
Nachschrift. Nach Ablieferung des Manuskripts stie.s.s ich auf
E. Benvenistes Aufsatz Tit res iraniens en armenien, in Rrv. ef. arm-.
9 (1929) 5-10, der sich an zwei Stellen mit nieiner Darlegung beriihrt.
Den armenischen Furstentitel isxein leitet Benveniste von mir.
*xsan ab, das er neben das oben S. 739 behandelte sgd. xswn, xsywn
stellt. Arm. sepith will auch er. wie Marr und Andreas, mit parth.
vispuhr verbinden, jedoch so, dass er es Uber *vscapurh auf air.
*visya-pu9ra zuriickfuhrt. Auch diese Erklarung ist unannehmbar.
solange nicht der Verlust des anlautenden c durch Parallelen glaubhaft
gemacht ist.
On the “ Uralian ” Element in the Dravida and the
Munda Languages
By F. Otto Schrader
I T will be remembered that sixty years ago Bishop Caldwell i
expressed the opinion “ that the Dra vidian languages occupy a
position of their own between the languages of the Indo-European
family and those of the Turanian or Scythian group ”, and that,
whereas the particulars of agreement with the Indo-European family
pointed only to a '■ very indefinite as well as very remote ”, if at all
any “ real relationship ”, those with the ” Sc}'thian ” family, and
especially the Finnish-Ugrian languages, were so close and so numerous
that they naturally suggested ” the idea of a common descent The
hj-pothesis was declined by the official science of language, but I found
it to contain a kernel of truth and thus wrote in 1924 my study
Dmvidinch und Uralisch,- intended to show that there was, if not
relationship, at any rate a most remarkable, though but partial,
agreement due. presumably, to the intercourse or mingling of two
nations in prehistoric times. I have since left the problem alone,
but feel now induced by Wilhelm von Hevesy's book Finnisch-
Ugrisches aus Indien (Wien, 1932) and his paper Feue Finnisch-
Ugritiche Sprcichen (Die Muiidaspracheit Indirns) ^ to take it up
once more. My special aim this time is to meet an adverse criticism
concerning word-equations, and I shall conclude with a few remarks
on the extent, kind, and probable provenience of the Uralian element
in both the Dravida and the Munda languages.*
^ A Comparatirf Gramjnar of the DroiiiUan or i^outk-Indian FamUy of Languages^
third edition (reprint of >econd), London, ltU3, })p. ix tf.
- »See Ziit-srh/ iff fur J/nloIogle und Intnisiik^ vol. iii. pp. 81-112. Of the additions
I could make to that paper one at least, \iz. to the paragraph on nuiner.ils. will be
found interesting : Tamil pnttu, Teingn etc., ten " is evidently = Samoyede
het, hu<l^ hii. etc. ; and Tamil, Tehigu (etc.) nufu " hundred “ seem?' to correspond
with Samoyede (Arch.) jur do. {< *h>iru : cf. Tam, ndl " day " with Samoy. jnJia,
jnh\ etc., do.). (Cf. O. Donner, Siuiinpdt''i'hf W OrUrvfizckhni.s.'-f, rejirint of 1932.
pp. 3.>. 9, 8. 48, 97.)
^ Pu])li''hed m 193.7. in the Atti dtl < 'ongre.-f-nt di Lingutstira fniuto in Eonm.
* 1 u>e the term Uralian (= Finno-Ugrian with Samoyede. or ancestor of both)
m an indetinite ''en>e. both for the <ake of commo<lity an<l aL^o to a\oid pronouncing
an opinion as to the special form or forms of speech with which we are concerned.
W. V. Hevesy pleads for the Ugrian, Caldwell was inclined to gi\e Finnish the
752
F. O. SCHRADER
Essential agreement of the grammatical systems is the conditio
sine qua non of the relationship of languages as understood to-day.
But to what extent grammar may change, even within a comparatively
short time, we know from the Romance languages and such Indian
languages as Hindi, Bengali, etc., as compared, respectively, with
Latin and Sanskrit. In the ca.se of very remote relationships grammar
to the extent demanded by our science of language for proving relation-
ship may no longer he available. Surely there are very many distant
relationships for which that criterion cannot be applied. IV e must
in those cases rest satisfied with stating connections between the
languages as shown by their vocabularies with or without such scanty
renmants of grammar which in themselves would appear to be
insufficient for the proof of relationship. This is what I have tried to
do in my study mentioned above. “ VV’hen it becomes clear, ’ says
Professor Setiila,^ “ that there are common words in languages which
have had no connections for centuries and even millennia, this very
fact is a valuable gain not only for the history of the languages but also
for the history of men.” It is true that related languages are referred
to here, but is not the sajdng equally applicable to the related
vocabularies of languages otherwise unrelated or of as yet unproved
grammatical relationship ? Yes, the critic would reply, provided the
comparing could be done in a strictly scientific way, as indicated by
Professor Turner, who in his criticism of v. Hevesy's book -, objects
that “ any correspondences, whether of grammatical structure or of
vocabulary, must be shown to be part of a regular system ”. Now to
this sentence, of course, no true linguist will refuse to subscribe. But
here is a demand which in those many cases of distant and obscmed
relationship can be, if at all, but very imperfectly obeyed, and which
at any rate cannot be complied with from the beginning. The phonetic
preference, but we may have to go farther back and perhaps less far for the Munda
than for the Dravida family. The following abbreviations will be met with : Ka.
— Kanarese, Ma. = Malayalam, Ta. — Tamil, Te. ~ Telugu, Tu. — Tulu, f'a.
=Santali, Mu. = Mundari, Cher. = Cheremisse, Esth. = Esthonian, Fi. = Finnish,
Hu. =: Hungarian (Magi^ar), Lp. = Lappish, Li. = Livish, Mord. = Mordvinian,
Ost. = Ostiak, Perm. = Permian, Syr. = Syryenian, Veps. = Vepsish, Vog. = \ogul,
Vot. = Votiak, Samoy. = Samoyede.
^ I translate from his excellent pa|)er Zur fruge narh der vprwandtt>chnff der
finnitich-ugnsrhfn und samojpdischen Sprarhen (Helsinki, 1915), p. 26.
^ See JBAS., 1934, pp. 798 ff. This is a rejection of v. H.’s attempt as is also
Sauvageot's review in BSL., 1932, pp. 180 ff., while Figulla {OLZ.y 1934, col. 187 ff-)
contends merely that the Munda languages belong to, but not that they are connected
with the Finno-Ugrian family.
“xjkaman " element in the dravida and mltnda languages 753
laws must first be found, and they cannot be found without lists of
tentative word-equations.
There is, however, in the world of linguists a great prejudice
against such lists in all cases where grammatical relationship is not
already probable. Wherever this probability is missing or but very
vague or no such relationship is claimed our h^-percritical linguists
would resort to the great Jlaya of casual resemblance and pity the
authors of the lists for the precious time they have lost. E. Levy
has declared ^ that such lists as mine (of sixty equations) may easily
be drawn up for any two languages, and has given as an example
sixty Kechua and Finnish word-equations. This is, however, not,
as the author of the Korean-Indogermanic hj-pothesis has (to
strengthen his own case) approvingly called it, a successful parody,
but a palpable failure. The list (which, to convince, should have been
one of proven cases of accidental resemblance) is marred by its
tendency ^ and shows, if anj-thing, that there may be ocean-spanning
linguistic relations like those assumed by' such scholars as Trombetti
and Rivet. But I admit the usefulne.ss of safeguards against chance
and shall now try’ to give some with examples from Dravida Uralian.
The ideal means for calculating the part played by the caprice of
language would be to search the whole material of well-known languages
for word-equations which can be proved to be illusory. Lists won in
this manner do not exist as yet, so far as I know, and will not easily'
be forthcoming. There are, however, two simple criteria for the
improbability of chance in individual cases.
There is, first, the test suggested by Meillet and used by v. Hevesy'
in his above-mentioned paper for a list of twenty-four equations (out
of forty -two contained in the typewritten copy distributed immediately
after the congre.ss) of Finnish-t^grian and Munda words all of which,
“ while designating quite simple concepts show three consonants in the
same order. " How safe, relatively, this criterion is can be gathered
from Enno Littmann's interesting paper Sprachliche Seltsamkeiten
aus Morgenland und Abendland {ZDMG., 1922. pp. 270 flf.). where
^ In his review in Zeitsckr. f. vgl. Sprachforschung, 1928, pp. 145 ff., written after
three scholars (P. W. Schmidt, W. Porzig, and H. Guntert) had pronounced, though
with a few words only, in favour of my attempt.
^ Xobody will accept such equations as L.’s Xos. 2, 8, 11, 14, 19, 23, 36-9, 44, 53,
57. They show his difficulty to reach the number sixty, whUe my paper has really
about a hundred equations. It is not the same thing to draw up lists purposely based
on superficial resemblance and such where etj'niologically impossible equations
are sought to be excluded.
754
F. O. SCHRADER
among seventy-two cases of probable or proven fortuitousness of
resemblance or identity (collected and discussed by the author)
there are only two which conform to the said condition (viz. Kechua
huarmi woman ~ lliddle and New Arabic liurma do., and Telugu
'pampu to send ~ Greek Trc/iTrco). H.'s list contains words of even three
syllables which are almost perfectly the same in the two languages
(such as Sant, cucunguf to sit on the heels ~ Hung, csiicsunget-
do.), and of such there is none in L.'s list and, though we may supply
two or three, such as Sanskrit upadhydya ~ Finnish opettaja (both
= “ teacher ”, but with entirely diiferent eti^miology),^ they can never
remove the impression that H.’s list alone is sufficient to prove the
connection between the Munda and the Finnish-Ugrian vocabularies.
For the Dravidian, however, with less direct Uralian relations, the
three consonant criterion will be found to be on the whole inapplicable,
such cases as Tam. vilahgti to shine ~ Hung, vildg light (and those
to be mentioned in the rest of this paper) being too rare to prove much.
The other criterion is taught by Heinrich Koppelmann in his book
Die Eurasische Sprachfatnilie, in an admirable chapter (pp. 13-21)
on the problem of chance in comparative philology. His thesis comes
to this (p. 17), that there is every likelihood of real (not illusory)
correspondence, when several connected, i.e. not arbitrarily selected
elements, such as the two negations of Indo-European and Korean,
are both found in the two languages compared. He further refers for
this to the numerals (1. 2, 3), pronouns (Fi. wd, fi, n-sa ~Lat. me,
te, se), case suffixes, and other parts of the grammar and word-
formation.
Now this means is applicable on a much larger scale than
Koppelmann seems to have thouglit of, viz. to practically the whole
of the vocabulary. For, it is evident that the admissibility of any
word-equation is heightened, the probability of chance lessened, if
another equation can be coupled with it on the strength of its meaning
or form or both, the likeliness of chance being reduced with every
additional equation. The connecting link may (1) be one and the same
meaning, as in the couple of equations meaning ' dust ”, viz. Ta.
podi ~ Fi. pdli -f- Ta. liimhu — Fi. tomu ; or in the one referring
to ■■ .shining ”, viz. Ta. rilaiigu to shine {vilakkn lamp) ~ Hu. nk'ig
^ Does Sanskrit pippllikn Votiak sippelika-'i (both = “ ant ”) also belong to
this category V A Hindu was delighted to Hnd in Germany the good Indian name
Kistenmac her (Kri&hnamacharya). Ofwordswith but two consonant?', but missing in
Littmann s list, at lea.Nt English mveh (ef. Gothic -^Spanish mucho {< Latin
multifi) deserves mentioning
“ URALIAN ■’ ELEMENT IX THE DRAVIDA AND MUNDA LAXGUAGES 755
light + Ta. nagu to shine ^ Fi. nakyd to appear, become visible,
or in the equation of words meaning “ small, little ", viz. Ta. cifu
~Hu. csiri + Ka. Z;iV/u~Hu. licsiny + Ma. piece ‘ — ^ Hu. pieziny
+ Te. pisaru ~ Fi. pisku. Or we may (2) have likene.ss of form
with difference of meanin;:. as in Ta. miyu to .sliine, etc. (see above)
+ Ta. nagu (nagai) to laugh ~ Fi. nakra do. ; or in Ta. cem red
~ Perm, cim do. + Ta. cetn straight, regular, etc. ^ Perm, cim
perfectness ; or in Ta. narnl to .sound, roar, etc. — Fi. nnrci to creak,
rattle + Ta. yiari to deride ~ S\t. nera do. + Ta. net grow thin,
emaciate ~ S\T. nermi do. (+ Ta. uarangu become blunted ~ Lp.
ndlu do., Fi. ndldkkd blunt instrument). Parallels in derivation
seem also, at least partly, to point to a common origin, as e.g. in Ta.
kumi to heap, Ka. kianbti bending ~ Fi. kiinw curvature + Ta.
kumbal a heap ~ Fi. kumpu hillock + Ta. koppul a bubble ~ Fi.
kuplu, Lp. koppak'k do. + Ta. kiimir (kitmili) to bubble forth ~ Fi.
kummuta do.; etc. There are, finally, (3) the cases where a class concept
adds strength to a couple or series of eejuations. An example (where
as will be noticed, as also above with koppul, etc., Meillefs demand
is also fulfilled) is the following one from the class of animal names :
Ta. nintiru bee (humming insect, from Hiiuir to hum) Samoy.
nlberii, nimere gnat + Ta. uanMli dog (prob. from netnal < *hamal
to vagabond) ~ Lp. njoainmel hare.^ But, of course, the wider the
class concept the less support can it lend to this sort of connected
ec^uations, unless they are numerous enough. I trust, however, that
the two larger lists I now give (with 3Iunda correspondences in the
second place where I could find such) will go far to serve the purpose.-
^ Here the rarene.'^s, in J^ravuhan, of the initial palatal iia>al a jK)int tu be
noticed (it has been mostly replaced by the dental nasal or /. •/. or has altogether
di.sap]x\ired ; ef. the nom. sing, of th(‘ Hr'^t jXTsonal pnmoun : Ta. yr'in (lit.), van ;
Te. nenu, enu ; Ka. nann, itnu, etc., all from pre^'Crved in Ma. only, uhith must
have been an ancient non-literary form also in Ta., etc,), as is the fact that both words
so far as known, do not occur in the '‘jdiere of intiueme of the Muiula languages. As
to the well-known vague use c>f animal nanle^ I may note from India that in Tamil
nan " jackal " (i.e. •• baw ler **) i.s a name of the tiiier also, and that in Kurku the dog is
called rhita which is in reality a kind of leopard.
‘ lioing l)iit quite imperfectlv acquainted with Finno-Ugrian linguistics I have
liad to reiK^unce aiming at a con>.i.'«tent phonetical rendering of the words adduced
from that quarter (as 1 have aUo simply tran‘«*literated in the usual wa\ the Dravida
and Mun<la word"), but have (a." above) emleavoured to give the Finni."h and
Hungarian words m their modorn orthography and have otherwise followed (with
simjiliHeil spelling) Setala and Szinnyei so far us they were availablr to me and. where
not. O. l)onner and occasionally v. Hevesy. The lists are also certainly not as complete
as they could be. Note that j in Fgr. and Ninioy. words is identic.d with the y in
J)rav. words (and in Engl. yutr. ete.) and that the n of the former is the same as
Drav. n.
756
F. O. SCHRADER
Parts of the Body
Hair, feather (wing, arm) :
1. Ta. Ka. navir man's hair ; cf. Te. navuru tender, soft, as grass,
hair, etc. • — ■ Fi. nava, naava beard-moss, ndnvelit moustache.
Cf. also Samoy. iiabta} abde, opt, etc. ; Ost. updt, Lp. vubpta_
etc., and (?) Fi. hapsi hair ~ Sa. up , Mu. up hair.
2. Ta. leaf mi collection, as of hair ; bundle, as of straw ; > kettai
hair ; Ka. Te. gafi feather, wing ^ Fi. kar-va hair, straw ;
Li. kar-a, kor-a hair, feather.
3. Ta. togai anything that hangs down : feather, plumage, woman s
hair {tuiigii, tukku to hang) ~ Ost. togel feather, wing : Fi.
tukka hair.
4. Ta. tu\n, tural feather ; Ka. tuppul feather, plumage, sott
hair ~ SyT. tyv feather, wing ; Vot. tdwel wing.
5. Tu. tuyi feather ~ Sa. duci tuft of feathers on head of bird
^ Samoy. tuja hair ; Lp. soagje wing, sleeve, Wot. suj arm.
6. Ka. pili a peacock’s tail, feather of such ~ Hu. pelyh flock of
wool, etc.), down-feather.
Lock :
7. Ta. puri (also verb : to be twisted, to curl) ~ Sa. phera forelock
Wot. per, pur, Hu. fiirt lock.
8. Ta. curt, curuttai hair-curl, curiyal locks of hair (from curi, curi
to curl), Malto curki lock of hair ~ Fi. suortua hair -locks,
Hu. szbr hair (cf. also Sa. carhi mane ~ Cher, sar horse-
hair).
Moustache :
9. Ta. rnicai, Te. nmsamu ~ Sa. misi ~ Fi. vlkset (Cher, miz wool).
Eyelid, eyelash :
10. Te. r'eppa, Ka. reppe, Ta. reppai eyelid ~ Fi. ripsi eyelash,
fringe. Cf. Sa. riprif to wink with both eyes.
Earlap, ear ;
11. Ka. pale, Jialle earlap ~ Ost. pol, Ha. fiil ear.
Mouth, lip :
12. Gondl torda, fork mouth, Malto torotb do. ~ Savara (oden
mouth ~ Fi. tursa, turpa snout, trunk.
(2) Malto tebo ( < *terbo) lips < — ' Syr. Perm, tirp lip.
13. Ka. cunca a bird's beak ^ Hu. csucs snout, beak.
^ For original guttural nasal (lost or replaced by r), see Szinnyei ; for Drav.,
cf. n < n below No. 50 and, above, la.st footnote but one.
“ URALIAN ” ELEMENT IN THE DRAVTOA AND MUNDA LANGUAGES 757
Lip :
14. Te. pedavi, Tu. bimmu ( < *bidamu) ~ Vog. pit’ mi, Samoy.
palmn.
Cheek ;
15. Te. hugga, Ka. huggi ~ Hu. bufa, pof{a).
Tongue :
16. Te. ndlike, ndluka ; Ta. ndkku, ndi'u ~ Hu. nyelv, Lp. njalbme ;
Samoy. ndmi. (For Drav. n < n see above, footnote on No. 1.)
Tooth ;
17. Ta. eyifu (egifu) tooth, tusk ~ Hu. agyar tusk, fang.
Jaw, chin :
18. Ta. alagu jaw ~ Hu. dll, Mord. ztl, Lp. ualul, etc., jaw.
19. Ka. avudu jaw, lower hp ; Ta. amudu lower lip ~ Samoy.
(Mot.) au'ol chin.
Neck, throat ;
20. Ta. kural throat, Ka. koral (kol) do. ; Tu. kekki (<*kerki),
kekkil neck, throat ~Fi. hirkku throat, Esth. kori do., Mord.
ki'rga neck.
21. Ta. cavadi-y-elumbu collar-bone ~ Lp. cdpat nec\^, Ost. sdbet
neck, back.
22. Ka. goiike throat ~ Hu. guga goiter.
Nipple, teat, etc. :
23. Ta. 7iagil woman’s breast, iVIalto naqlu uvula ~ Vot. naka plug,
peg ; Syr. nak do., a nipple.
24. Tu. nalli a spool ~ Fi. nalikka plug, nolkki membrum virile.
25. Ka. keujelu, keccalu udder ~Hu. csecs breast, nipple.
Limg :
26. Te. upiri-titti lung (‘' breathing-bag ”) ~Fi. fdtu, Hu. tiidd lung.
Armpit :
27. Ka. kauuii-kur, kahkur, Tu. ~ Mord. kaval-al, Esth.
kangla ; Samoy. kaiigla, kahgel, etc.
Back :
28. Gondi murcktd ~ Samoy. mdchal, mdgal, etc.
29. Hump, hunch : Ta. kun. kunal ~Esth. konar.
Breast, hip :
30. Ma. kohka, Ta. kongai female breast ~ Fi. konkka hip (cf.
Esth. kiinge hill ~ Ta. koiigu name of a hilly country).
758
F. O. SCHKADER —
Belly, stomach :
31. Ka. hasif, Ta. vayifu ^ Sa. bhosam ~ Yog. vas (Hu. has).
32. Ka. poUe, Te. potta ~ Sa. j)ota {potra big-bellied) ~ Hu. potroh
paunch, pot-belly.
33. Ta. poccai, Te. bojja paunch-belly ~ Hu. puczor stomach, paczo
and poczak paunch-belly.
Navel :
34. Ta. pokkid — Sa. buka ~ Fi. puka (Hu. pokla umbilical cord).
35. Tu. puvalu ~ Hu. pup.
Pudenda :
36. Male organ : Ta. Ka. mdni ^ Samoy. mane ; Lp. marine,
tnonne (egg). Hu. many.
37. Ka. Te. yolli female organ, Tu. kullige anus, Brahul gilhk do.
~ Fi. kuUi male organ, Esth. koli. Lp. guolla. Samoy. hyle
(testicle).
38. Anus : Ta. kunti, Te. kutte, Tu. kudi ~ Sa. kuthi ~ Cher, kotan.
39. Ta. poccu anus, Ka. pucci female organ ~ Lp. buocca male organ.
Hu. puczii, fast do. Mu. peso glans penis, Sa. peska testicles
~ Samoy. pije female organ ; Lp. picc. Cher, jns male organ.
40. Anus (■■ posteriors ’') : Ka. pifa, pefa ~ Fi. pern.
41 . Testicle : Tu. ari ~ Hu. here.
Hip :
42. Ta. eappai, Ka. eappe ~ Hu. ezipb, czipe.
43. Ka. tonka ~Y'\. lonkka. Cf. Ka. dot'iku to bend, be crooked
Fi. hjnkkcl crooked.
Calf of the leg, ball :
44. Te. pikka calf of the leg ~ Fi. piikia ball of the thumb. Esth.
piikk ball at hand and foot.
Finger, toe :
45. Ta. viral, Ka. beral, Te. vrelu finger, toe ~ Fi. varvas toe.
Nail, claw, hoof ;
46. Ta. iigir nail, claw, Ka. ugur, Te. goru ~ Hu. kdrbm nail, hoof.
47. Ta. kavai-y-adi (“cloven foot’"; cf. kappii, kavar forked
branch) ~ Fi. kavio hoof, Esth. kabi.
Bone :
48. Te. elurnbu, Ka. Te. elu ~ Fi. luu.
Knuckle, joint :
49. Ta. po-ruttu joint ^(1) Hu. porcz cartilage.
“ UEALIAN ” ELEMENT IN THE DRAVIDA AND MUNDA LANGUAGES 759
Sinew, muscle :
50. Ma. Tmramhu. Ka. nara (Ta. nar fibre, string) ~ Samoy. ner :
Yog. nciri, Lp. njorga cartilage.
51. Ta. ecam (sp. iesam) ~ Fi. jasen, Hu. izom.
52. Ta. Tnai nerve of a leaf ^ Hu. In. in sinew.i
Skin, hide, fur :
53. Ta. porvai covering, skin ~ Hu. 6o>skin, leather. Cf. " covering,
cover : Te. bimda ~ Hu. bnrolc.
54. Ka. togal, toval skin, leather, Ka. Ta. tol, Te. tohi, tokhi ~ Yog.
tmvel, tout fur, leather, Fi. taija pcllis pilosa.
Dirt and Excretions
1. Te. ro/nta, Ka. rote dirt, filth -^Sa. rotha black and dirty ~Hu.
rond dirty.
2. Te. rompi mud, mire ; ronipa catarrh, snottiness ~ Fi. rnpa
filth, mud ; Esth. ropnsta to soil.
3. Tu. rebkuni to be smeared, soiled, Te. regnda clay ~ Fi. rtikd.
Yeps. rega snot ; Esth. rdgase slimy, filthy. Lp. rnncca.sa
muddy.
4. Ka. masi, mdsalit. Te. /R/i.si dirt ; Ta. inUcn spot, stain, blackness
~ Sa. mxisra dusty, dirty Hu. niaszaf, nmczat dirt ; Fi.
niusta black, dark.
5. Ka. pisarxi, piccn filth of the body, rheum of the eye ~ Hu.
piszok filth, dirt, pocsek dirty, soiled.
6. Ka. sari, Te. cctri paste, gum. etc. ; Ta. cefti. cifumigam mud,
mire, curi become muddy ~ Mu. sara-gara dung-pit ~ Hu.
szar, Cher, sor mud ; Hu. csiriz glue ; Hu. szirtos, szartos
dirty.
7. Ka. Te. laddi, Ta. latti, Kurukli Uddi dung of horses, elephants.
etc. ~ Sa. Jed. Mu. Jed', lid', do. ~ Fi. Janta dung. (Cf. Sanskrit
landa, lenda, Prakrit laddi.)
8. Te. retta dung of birds, etc. Sa. res dregs, etc., ritha dirty ~ Fi.
reto, roto dirt, Hu. redves rotten, putrid.
9. Ta. kacadu stain, filth, Ka. kasaru dust, impurities, kisaru
a. gJju rheum of the eye ~ Sa. kacar, kacor dirty, miry,
Kurku kass dirt. Mu. gij-gij' filthy, kasrd itches ~ Hu. kcsz,
Esth. kds-n, Yot. giz-lo scab, scurf.
^ With h)ss, on iioth side'i. of the initial sibilant which is preserved in the Votiak,
Cheremisse, and Finnish ^ord (-'ew. For the Dravidian, cf., e.g., the word
for eyelash : Ma. cima a. una, Tu. a. imey Ta. only imoi.
760
F. O. SCHRADER
10. Te. pusi slimy impurities of tie eyes ^ Hu. fos thin faeces
(animal and Human).
11. Ka. lodale phlegm, spittle ~Sa. lodhor lodhor thick and gummy
(as pus) '~(?) Hu. latyak puddle, morass, Iocs, morass, lucsok
mud.
12. Ta. korai phelgm, spittle ~ Fi. kuola mucus of the nose.
13. Tu. noli phelgm, spittle ~ Hu. nyal, Fi. ndlje, Mord. nolgo mucus,
phelgm.
14. Ta. cali thick phelgm ; Ka. jollu, Tu. jolU, dolli saliva ^ Fi.
siilki, Esth. siilge, Lp. coly saliva.
15. Ka. Tu. ki-vu, Ta. ci pus, matter ; Ka. kl to become putrid
' — ' Mord. si, Lp. sai pus ; Ost. sij-ta to become putrid.
16. Ta. pina, Ka. pena, Te. jnnuge corpse, carcass ~ Hu. penesz
mould, mildew, penyved- to rot, putrify.
These lists are not the result of a systematic search of the
vocabularies concerned, excepting only that of the Kanarese. I cannot,
therefore, except for the latter, give ciphers for calculating the
percentage of words included in the lists. In the Kanarese (Kittel's
Dictionary) I have coimted 84 words (not including trifling dialectical
varieties) for parts of the body, and of these there are, i.e. could be
(some are omitted for brevity’s sake), just 30 in our list, i.e.
36 per cent. For the second list the corresponding numbers are 29 and
9, i.e. 31 per cent. For v. Hevesy’s equations the result appears to
be less favourable. I have counted only the words and equations for
parts of the body and found that there are 136 such words in Campbell s
Santdli-Enylish Dictionary and 45 Santall-Uralian equations for parts
of the body in H.’s book. There are, however, among the latter about
a dozen which I should have eschewed, and I am perfectly sure that
my list is less complete than his.
It is interesting to see how far the twelve parts of the body (hand,
foot, nose, eye, mouth, tooth, ear, hair, head, tongue, belly, back)
for which the Linguistic Survey of India gives “ standard words ”
are represented in our Dra\'ida-UraUan list and among v. Hevesy's
Munda-Uralian equations. It will be seen that of those parts seven
figure in our list and (discounting one or two quite vague identifications)
five in H.’s book. As to the “ standard words ” given for those parts
from the several languages — about forty for either family (cotmting
related words as one) — there are only three of them in our list {torotji,
etc., vayifu, etc., murchul) and six in H.'s book (khuri, kata, etc..
“ URALIAN ” ELEMENT IN THE DRAVIDA AND MUTOA LANGUAGES 761
mehge, up, etc., pehg, putti = pota) and, moreover, two doubtful
items on either side, viz. (1) the commonest Dravidian word for
“ hand ” (Ma. I’ogyi, Ta. I'ei, etc. ~ (?) Fi. kiite, Syr. ki, etc.), where
the guttural of Kui kagu (by the side of kaju) and KurulA Mekkha
seems to stand in the way ^ ; and (2) the widely spread ilunda word
meaning also “ hand ” {ti, iti, etc.), where the vowel casts doubt on
the comiection with Sanioy. ulte, ude, etc. (the word is not in Finno-
Ugrian). For the possible connection of Sant, up' with the Dravidian
see No. 1, for pota No. 32 of our list ; the remaining words have no
correspondence in the other faniilv.
By far the largest part of the vocabulary both of the Munda and the
Dravidian family is. at any rate, no! Uralian. The imposing number
of 1,134 word-equations given by v. Hevesy must not deceive us. For
Campbell's Santdll-EngJish Dictionary, on wbich they are based,
contains some 20,000 words, and the.se can hardly be cut down to les.s
than 5,000 non-Sanskritic word-stems. And though not so many as
every second or third of v. Heve.sy's equations will have to be cancelled
(as Figulla seems to believe), the increase through fresh ones will
barely suffiice to make up for the loss. For, admitting that our know-
ledge of the Munda languages is as yet but imperfect, we still know
this much that most of them, viz, the Kherwari group (comprising
88 per cent of the speakers of Munda languages), are really but dialects
of one and the same language with practically the same vocabulary,
whereas the small rest (Kiirku, 8avara, etc.) is characterized by over-
whelming foreign influences. In the Dravidian vocabulary also the
Uralian constituent will hardly be found to e.vceed one-tifth part of it.
There remains the question as to the kind of connection between
the Uralian element of the Dravidian and that of the Munda language.s.
W. v. Heve.sy cuts the Gordian knot by simply denying " that the
Dravidian languages (apart from borrowings of which, indeed, there
are many) have anything in common with the Finno-Ugrian language
family ' (OLZ.. 1934. col. 478). Now I admit that the grammatical
correspondences pointed out by me ten years ago are not sufficient
to a.s.sign the Dravidian to the Finno-Ugrian familv (nor were thev
meant in this sen.se. but only as vestiges of a more remote " Uralian '
affinity), whereas those shown liy v. Hevesy do seem to allow us to look
at the iMunda family as the denaturalized offspring of a lost child of
the Ugrian branch (or primitive Ugrian). But it seems to me quite
^ Unless some such evolution .'is kaifi > kairn > kayyi > knju > kagu and
. . . knyyi > kf.yye > khfga > khekkha could have take place ; cf. equation Xo 2.'^.
VmL \HII. parts 2 AND 3. 41)
762 “ URALIAX ’’ ELEMENT IN THE DRAVIDA AND MUNDA LANGUAGES
impossible to regard everything Uralian in the Dravida languages
as borrowed from the Munda ones. Such borrowings have, of course,
taken place (as is obvious, e.g. in the case of words beginning with
initial r and /), and there may be even a Munda " substratum ” in some
Dravidian languages (such as Telugu, as Professor Bloch is inclined
to believe). But then the Munda languages have borrowed as much
(if not more) from Dravidian, and Tamil and Malayalam at least have
no Munda loanwords (except a few taken ov'er from Telugu or
Kanarese), nor do they look like having a Munda substratum, which
is, indeed, not to be expected in those southernmost countries of
India, where Munda tribes have never been, so far as our knowledge
goes.
I therefore come back to my hypothesis ^ that primitive Dravidian
(proto-Dra vidian) entered India from the north-west with Lrahan
elements already assimilated, and consequently assume that the
Uralian element of the Munda family is not directly connected with
the Dravidian one, but points to a marriage between an indigenous
Indian tongue (hailing probably from Further India) with an Uralian
invader who thus became, possibly without having come in contact
with the Dravk.las, the dominant agent in the formation of what are
now the Munda languages. The two other hc^iotheses which appear
to be possible here are (2) that an Uralian invasion has directly
affected, but in different ways, two indigenous families of speech ;
and (3) that there have been two such invasions each of which has left
its trace in one of tlie two families concerned. In no case, of course,
can we expect to find imported elements only.
^ Suggested by the geographical position of Bnihui («ee on this now
G. .Morcenstieme in his Report on a Lin^viMic Mission to Sorth-\\ eslerii India, O^h).
1932, pp. 5—6), but necessitated also, I believe, by the pre-Aryan “ Mediterranean
component established for pre-historic India by anthropology and archatology and
equally required by certain linguistic considerations on which, however, I am not
now prepared to speak.
The Numerals in the Niya Inscriptions
By 0. Stein'
A. Cardinalia
1 FOR “ one ” the NI use eka (106 Uo 7. 159 r 2. 187 o 8. 211 r 5.
231 r 2. 248 o 8. 253 r 1. 348 Uo 4. 425 Uo 5. Cr 6. 437 Cr 1. 577 Uo 6.
630 o 4. 638 o 2. 3. 676 Uo 5. 688 oA 3)4 In the form ega it appears
709 Cr 4. It is used in the uninflectod form, irrespectiv^e of gender ;
the genit. sing, ekisya “ single is found 272 o 10. 514 o 1 ; a plural
f'ke occurs 468 Cr 2. eyaya in 417 o 2, as nomin. masc. ryago in 296 Uo 3,
corresponds to Skt. ekaika-, Pali ekaka- “single", in accordance with
the phonetic value of -ga.- eka~ is found in compounds with numerals
(see 11), nouns (ekadesayvni 272 o 6),® adjectives {ekavarmga =
ekavarsika- 589 Uo 3). In Skt. 511 o 2 occurs ekdbhirdnm-, rl
[ekamanasa-] ; an amredita ekameka- is found 109 r 5. 140 Cr 6.*
2. The numeral for “ two ” appears in many forms without regard
to gender and case, and even the number is ignored in most of the
passages. As dvi it occurs 72 rB 2. 5. C 2-5. D 1. 2, 4, 6, 7. oA 1, 2,
6. 7. B 4-7. D 4. 7. 198 o 2. 546 o 2. 570 Uo 5. 665 o 6. sometimes
written tvi (81 rE 3. 719 o 2) ; dvii 100 r 4 is perhaps drili “ for the
second time judging from the connection with ayogena, appearing
in 539 in a similar phrase ; aijog[e]na huda bhiti “ without gain,
profit’’; cf. tr[e']ya, dvaya ayoge 100 r 3, 5. drim (577 Co 1, Uo 5)
seems to be a peculiarity of this inscription, either the Anusvara
indicates long 7 “ or the form may represent a genit. plur.. recte dualis,
for *dvina, according to tri)ia and the dependence on praceya, con-
structed with the genitive, e.g. kkidana (40 Ur. 55 Ur) or 46 Ur
' Atibreviatiuns : A. B, C. etc. = Columns. C = Cover-tablet. NI = Kharo^thi
Inscriptions discovered by Sir Aurcl Stein in Chinese Turkestan, o = Obverse,
r = Reverse. U — Under-tablet.
" Kapson. Xl, p. 301.
® “On a single place ; Thomas, AOj xii, 40, €ka-^ti 416 o 5 between rnkora
and iiti seems to qualify the latter which is called asti also lS0oB2; again 272 o
a horse is called : it seems that also corn 272 o 6, 7, as ghridn =■ Skt. ghrtn
l.)Hr4; on the other hand, occurs in conditional •'cntences with ydti and the
verb fis “ to be Its meaning seems to be that of Skt. a-^tfyn “ extant " ; ekasti
“ single extant tlu'rf'fore xud itti 1. naMi is negative, cf. Thomas, AO., xii. 43, 7.
* On eka h[h)iti- *ee below.
Cf. Thomas, AO, xiii, oO, 2.
764
O. STEIN
(Konutnae).^ dii appears 571 Uo 3. 628 oA 4. 703 r 4 ; a contaminated
form dui, perhaps due to weakening of -e and found also in Apabhramsa.
occurs 21Uo2. 141 o 1. 180oB2, 4. 214 o 3. 215 rl. 291 o 4. 5.
364 r 4. 414 r 6. 520 Uo 3. 638 o 2, with the variation of writing tin
106 Uo 8. Cr 6. due is a real dual {miliml 422 o 5. pitaputre = °putrdh
715 o 2) for fem. and masc., used as plural. In compounds di- is
met as the first part in diguna {= dviguna 142 r 1. 283 o 2. 351 o 2.
387 o 2), while in connection with -vara = -vara the- form dvi- is
used, if not the ordinal form (see below B. 2).
ubhaya. As a dual-plural ubheya is found in 580 Uo 2 {te u. hhuma).
Cr 1. 581 Uo 3 ; iibJiayamdadc ® 387 o 2 would correspond to Skt.
vhhaydntaiah, the duty is to be performed two and three times at the
sacrifice of life (for the reward ?) in both the worlds,^ according to the
order, says the inscription pleonastic m this passage.
3. “Three” is in the NI tre (72rBl,6. C 6. D 3, 5. oA3. 4.
B 2, 3. C 1-7. D 1-3. 6. 383 r 6. 415 Uo 3. 715 o 3) ; treya (100 r 3 i
106 Uo 6. 157 r 1. 162 Uo 5. 571 Uo 5. 593 Co 2. Uo 2. 630 o 2. 721 o 5).
Inflected forms are found 431 Uo 1. 432 Uo 1 : trina sayivatsarana,
a genit. plur., which is probably also the case in 671 o 1 ; in the
mostly Skt. inscription 511 r 4 the three pitaka {pi'dakdni trim) are
mentioned. As the first part of compounds appears tri- in triguna
(283 o 2. 351 o 2. 387 o 2. 676 Uo 4). as in Skt. tribhava 511 r 5 ; in
adjectives like trevarsiya (593 Uo 5) besides treya varsi (593 Co 2.
Uo 2. 341 o 3, see n. 1 ibid.), trevarsaga (152 Uo 4).
4. “Four” is catu (27 Cr 1. 496 Uo 4) or cara (348 Lo 3) or
caura (70 Uo 3. 720 A 5. 562 Uo 2) ; it is tempting to take cohiira
637 o 6 (aniTta pasava 4 cohura miliaia amnehi nidae huamti) as the
numeral “four ", but neither numerals occur in that inscription
^ A similar termination -im occurs in 117 o E : mrvaphndfii/n fntn ('fi)n yavi 20 3,
which has been explained as Mirinni liinddM tdni ydroh 23 by Professor Thomas
{JRAS, 1927. 545) and as sarrdh pinditd yavah 23 bv Professor Luders {SPA U , 193d.
iOlO. 2) Avho prefers to read : pimdaitntamcam. Generally the XI use only sarnt
phndn^ so that a verb Skt. pindny- is less probable, also on acc'ount of the expression
eka pimda (24So8). wuh or p, viidi (.390 Uo 4. 428 o 5), or sarvapimda
yanainnPDfi )/iidi (345 Uo .3) \^h«’re it must be a noun. On the other hand, 587 lo3
offers nrauuta hhuma riknda which may be *cirat/itfi bhumir vikntd (land which
has not been tilled a long time ? has lieen sold). Then -imta would jxiint to the part.
j>erf. pass, of denominative verbs,
^ Instead of an. The pronominal termination -e of the jdiiral is found in khida
putro 415 Cr. 3. ynnmnn.'tt 139 I o 3. /y/.v#’ ,331 oB 2. rrdkp 329 r 3.
Cf. surya{s^rsa ?)rfade, 572 Cr.
* Cf. Kp. Ind., XX, p. 18, P 2, line 5, etc., ju 27 s.v. nhhaijfdokahiUp . In SI, 107 o 2
is mentioned in the addre-'S the alamlrtaparahiyamatya-.
NUMEKALS IN THE NIYA INSCRIPTIONS
765
besides the signs for them nor h is inserted between vowels to avoid
the hiatusd In compounds, beginning with “ four ”, either catu- is
used {catuvarsi 420 o 5 ; catuvarsaga 431 Cr 3. 432 Uo 8. 437 Cr 7) or
caura- (70 Uo 2 : cauravarsi).
5. pamca is the numeral for " five ” (180 oB 3. 187 o 6. 211 r 8.
261 r 1, 3. 327 o 5. 329 o 4 [genit. plur.]. 415 Cr 1. 437 Uo 2.
439 Uo 3. 506 o 9. 519 o 3. 630 o 4). The same form is met in com-
pounds with -varsaga (432 Uo 5), while 431 Uo 7, being apparently
the rough-copy of 432, shows paca- ; 591 Uo 3, 4. Both these forms
are used also in the additive numerals (see 15. 25).
6. The numeral for “ six ”, so, is found 439 Uo 3. 519 o 3. and
in the compounds sovarsi fern. (27 Uo 2. 676 Uo 3, 4), cf. 26.
7. sata appears 468 Uo 4 and in the compound satavarsaga
(209 0 3. 575 Cr 2) for seven
8. “Eight” is met as atha (with loc. varse.v() 256 o 3. 5. as in
the compound athavarsaga (341 o 3), athavarsi (625 o 6) or as asti
(661 0 3).2
9. For “ nine ” is found no 576 Co 2 and 715 o 4 in ?i[o]rarsa^a ;
cf. sub 90 and ordinalia sub 9.
10. As in Skt. dasa is used for “ten” (cf. Pischel, Pkt. Gr.,
§ 262, p. 182. for Mg. and Dh.) 222 o 5. 327 o 4. For dasavida and
similar forms see 100.
11. Only the ordinal numeral occurs, see below.
12. For “twelve” dvadasa is found 346 r 2, hadaha 419 Uo 4.
579 Uo 4, which latter stem occurs also in the ordinal.
13. Unlike as in Pkt. for “ thirteen ” a contracted form trodasa
appears 431 Cr 7. 432 Cr 4. 505 o 2 (ibid., n. 2 : tre°, cf. Plate VII).
15. In accordance with pamca and dasa “ fifteen ” is panicadasa
489 r 9.
16. sodasa is “ sixteen ” 514 o 2 ; [jo]dasa in 16 o 3 for which
in note 3 the alternative reading [/o]- is given, would point to sodasa.
This seems a rather long time, since no camels have been brought ;
that j represents also a cerebral sibilant may be due to the foregoing
ajhi = adhi “more than sixteen years”.®
20. “Twenty” is visati 275 o 2. 393 o 2. 576 Co 2. Uo 5 : for
vimsa see 26.
^ In Saka exists tcahaura “ four ”, Stcn Konow, Saka Studies, Vocabulary, p. 184.
- atn (358 o 6. 637 o 3) is not adn (Index s.v.), but, according to Thomas, A(cta)
0(rientalia), xiii, 67, Hindi dtd, Skt. atta “ flour ”. For asti cf. Noble, BSOS, vi, 453.
’ Cf. Burrow. BSOS, vii, 51 1 on ? > s. s> j>'), .i > jh(z), see also V/, p. 303.
766
O. STEIX
25. “Twenty-five” occurs 225 o 8 as pacavisa.
26. It is the peculiarity of the NI that phonetical rules to be
observed in some instances are soon contradicted bj a next instance ;
if visa represents a Pkt. visa, in sovimkt 207 r 3 the Skt. form appears
again ; though the Anusvara in visa is dropped, one may suppose
that the i is long.
27. Analogical “ thirty ’’ is expressed by trim (186 o 3. 387 o 10.
495 o 5. 590 Uo 3. 592 Uo 3. 609 r). the same word seems to be contained
in satrisa, explained in the Index, p. 375, as a compound with the
adverbial prefix sa = saha. satrisa occurs three times 209 o 3. r 1. 2,
followed by avimtama the meaning of which must be “ fine, punish-
ment ”.i But as the penalty is stated only in r 2. satrisa seems to
correspond to Skt. sadrsa “ suitable The Index, p. 350, explains
dri'ju 661 o 3 as Skt. trimsat, and also Noble takes the word in that
sense.^ aghita is no doubt some kind of official title, variations of
which are agita. ageta, argil a ; as in the many other passages where
the title occurs, a personal name is required, Driju Vaso must be the
name of the official.
40. Forty ” is caparisa (580 Uo 3, 589 Uo 3. 590 Uo 3) where
tv > p.
42. The same stem appears in dn caparisa “ forty-two " and
45. in paijica caparisa, " forty-five.”
50. The numeral for “fifty" is pamcasa (437 Cr 7. 567 Uo 3.
571 Uo 3. 676 Uo 6-7) ; here, as in caparisa- dsa. resp. -tsa are
e.xpressed by the short vowels.
70. For "seventy" satati is found 571 Cr 5-6. 580 Cr 2.
90. “ Ninety ” is novali 655 Uo 7.
100. In the numeral for “ hundred ” the dental shows plenty
of variations, sata (149 o 2. 225 r 3. 345 Uo 10. 668) is written with
the medial as sada : 368 o 4. 567 Uo 3. 740 o 3. This numeral has
also some inflected forms, as it seems, especially often occurring
^ Cf. Burrow, /AS'OaS', vii, 784.
^ 338 r 1 seems to contain [ifroyiatni na rc aiaia rncKn prnsaretu “ (whether)
there will he .-uch one in the drolla or not, at any event the VacKii must be sent away
(let free) lacHtt is fouiul again 030 o 2, 3 mth jnmnn ami the verb pra 3 nvita ; for
him fifteen men are to lie given, is he perhaps a man from Vak^u, the Oxus ? tor
drsta appears tritha. ff)r dr^tin in 510 Iritrrt,
^ BSOS, vi, 453. (Ihiihi is hanlly — tntkiiy as this word occurs o2 in the form
tntha ; like duhitr by metathesis of aspiration appears as tlhttu, dhaki would correspond
perhaps to dudhi ; beciiuse Apliinanu takes away the camel and the fujhitd I).V. the
rnilk, Khvarnase prefers to sell the camel. An objection against such an explanation
is the marsculiue sex of the camel ; one would expect uti.
NUMERALS IN THE XIYA IXSCRIPTIUNS
767
satade (82 o 1. 94rl[?]. 24. 29. 35, 43. 342 o A 2. 3. 701 rA 1. B 2.
C 4. D 1, 7. F 1. G 4. oA 1. C 1. E 1) or sadade (115 oA 1. C 1. El,
G 1. I [1], L 1. 132 rA 1. B 1. C 1. D 1. E 1. F 1. G 1. oA 1. B 1. C 1.
D 1. E 1. F 1. G 1. 342 oA 4. 650 rA 1. B 1. C 1. D 1. E 1. F 1). tin-
termination of which points to an ablative. The locative itilnninn
is found 46 Cr 2, 3. 73 oC 1. 74 oA 1. B 1. 4. rA 1. B 4. rA 1. B 4.
76 oA 2, 3-8. 92 o 1. 1. 1. 169 rA 1. B 1. C 1. oA 2. B 1. C 1. D 1. E 1.
342 oA 1. 5. 6. C 1. 2, or in the form sadatni (41 o 1|2 ]. 2 [3 > ]. 3.
rl. 168 o 2. 170 rl. 173 oA 1-7. 174 oAl. 185 oA 1. B 1. C 1. D 1.
221 o 1. 268 r 1. 299 o 1. 313 o B 1). The cohimn.s are mostly headed
by a personal name in the genitive and kit ade {sadade). after, i.e. beneath
that, follows a list of persons and their propriety of camels. In 132
the list is introduced by the date (year 30. month 9. day 5) and by
niu[sini]/ia )//an([w«n)] Uhitavya. madmm occurs 374 o 2 as locative :
masinammi samvatsari palpi ciinditaya. and 589Cr4f.. where the
text runs : (aha adha »mU Lpipimtsaae i/ida adha wuli ixnsina yatiwi
Cigitorijena gida. That seems to be ; '• half of the price received
Lpipinitsa and half of the price ‘ of the whole amount ' received the
yatma Cigitoya." In 374 the taguja Sudarsana and the gatma Acosa
are told to consider the annual ta.xes from the kibn-dga and from
the raja^ ; masinaiiuui, i.e. they may specify the two items " within
the whole amount ". If this is correct,- the headline in 132 means
“ document regarding the whole amount of camels " on the date
specified. The single hta (ktda) is in the hands of an individual, under
whom are people po.ssessing one camel. In 94 again a list is given
where from the sata {satade) the amount of corn and the camels of
individuals are specified. 115 offers a register of animals in different
sadas ; 650 and 701 the names of persons are enlisted. 169 has this
headline : In the year 26 of the regnal period of his exalted majesty,
of the devaputra Jifugha Maviri. in the 10th month, the people
have been made to measure their .vtki mma " ® ; the following is a list
of the measured khi s of mdividuals. each belonging to a sata of
persons ; .some of them, Gothaka. Lpipta, Ra.sena are on the head
^ On k th/irri and tajit, of. Tiioma^, AO, xiii. 63, re-'p. 4.> f.
- ma<nta- may t)c t“omuAt(Mi \^ith Iran, ma^au- ( Bartholomae. .4/fira/L Worterb..
1154), the adjective <ierivated from the noun in ma^an- hv haploloay in^tead of
*nia-^nna}>- ? HarthoKnnae refers to uirtnin-, udmifi-, and Wlutney, § l:23(V.
could he explained as n-stem or compared with the many proper names with suftix
■ utn, cf. Thomas, Fe<tgahf Jamhi, 62 f.
® Cf. .S7C4 ir. 1933. llKtl ; Thomas, .40, xni, 60. 2, and Burrow, BSOS,
Til, 510 f., accept the meaning wine
768
O. STEIX —
of more than one sata, as it seems. Such a registration is contained
in 180 where the royal camels are counted. With the suggested
sense ^ of sata, measure of land, the long hst of ra[c}iani\na jamna
“ guards in 701 is hardly reconcilable ; in 650 r sadade seems to range
with pradejami, no doubt corresponding to Skt. pradese, an adminis-
trative unit. That sense may fit sata (sada) and also the analogical
term dasa. 170 rl (cf. Fragment o 1) sadand occurs; Kapgeya is
on the top of it, under him there are two dasavida, one has to collect
12 Ichi of masu, the other 11 khi ; oA 1 mentions : asidha (ca-, ram-}
neya dasammi sesa, also under that dasa persons and the amount of
khi are enumerated ; further, 341 seems to mention the name of
a dasa in Pi-saliya.^ In Niya, therefore, existed some administrative
and fiscal units, pradesa, sata, and dasa, corresponding to similar
institutions in India. Then the titles of the heads of these units are
preserved in the NI. The Index explains s.v. sadavita “ denom. of
sahda ’’ in 159, 247, 683. From the phonology of the language in the
NI it is obvious that there can be hardly any difference between
tasavita (a dfl-af Xeyopevov) and dasavita which appears also as dasavida.
these two forms once occurring in the same inscription 204 ; thus
satavita in 683 may be only a clerical pecuharity.
To start with the inscriptions where sadavita is alleged to have
the sense of a denominative of sabda : 159 r 1-3 reads : —
adehi tusya mahatvana paride na kimci srunammi udisa. ahuno
sadavita Kolpisasya hastammi vacari 2 p-ahidemi. eka Sachammi
picatidavo biti Cadotatnnii gimnidavo . . .
“ Since then I hear nothing from Your Excellency with regard
(to that). Now I dispatched in hand of sadavita K. two vacarid
One has to be delivered in Sacha, the second is to be received in Cadota
(i.e. Niya) ”. 247 o 3 f. ; tnasu ahuno avasa sadavita Sugatasa hastammi
prahadavo, masu has at all events to be delivered in hand of a sadavita
S., sealed it has to be delivered, adds the text (nmmtritaga prahadavo),
specifying also the seal. 683 o 1 opens with : satavita patrana gorna
giltal . . ., as in the other lines a list of goma (cow-dung ?), small
cattle, and grain of seed or barley is given. 86 o 4 says : avasa
pratu sadavida Kdrsenade (-sa 1) tra)nghad}w.[;re] yo pravamnaga isa
garndavo, at all events the sadavida from Karsena who is able for
^ Thomas, SI Index, s.v, kaiadt^ p. 371.
® Thomas, Festgabe Jacobi, 48, suggests it to be Pialma.
® Cf. Thomas, AO, xiii, 56 f.
NUMEKALS IX THE KIYA INSCRIPTIONS
769
holding the office ^ early ^ has to go hither. Putting aside 422 o 9,
though also here katavida is followed by a personal name, in 436 Uo 2
the text runs : a^ti manusa sadaiita masdhige sacici . . . ; it looks
as if sadaiita would be “ a man called Masdhige ’ ; but Sacici is a
proper name as in 573 Uo 2, Masdhige his padrone who is mentioned
further on Uo 3. Cr 1 f., cf. the parallelism in Uo 2-3 ; sanmsena
karcikasa ca manusa rutrayasa. 482 Uo 2 relates of the complaint of
Saka : yatha edasa kilmeyammi molpina bhuma ladhaye sadavida
karsenaia achimnamti na o'demti krisatnnaye tade vrcYia chimnamti
vikrinarnti ... It is evident that the subject of the plural achimnamti
and odemti are the two persons, the sadaiita and the karsenaia, sc.
“ that the sadaiita and karsenaia do not take care ^ for obtaining land
on his (Saka’s) farm Molpina, that they do not deliver it (to him) for
tilling, further, that they cut trees and sell (them) ” ; it is not right,
continues the inscription, that some take away the property ^ of
others and sell it. As the karsenaia is some official, occurring many
times in the NI, the sadaiita must be also the holder of some office ;
noteworthy is the connection of that suggested officer with land- or
estate-affairs. In 715 o 4 f, it is said : puratkida mahatvana te hhuimm-
mi sadaiita Ricikgasa escari huda vavamnae kisivamnae amna prahuda
deyamnae “by Your Excellency (lit, in the presence of Y, E,) the
sadaiita R, has been granted the right on your land to sow and to
plough and to give others a present It must suffice to refer to the
many instances where the sataiita N.N, occurs with other officials
like cojhbo, gusura as witness : 569 Uo 7, 8. 580 Uo 5, 6, 586 Uo 6, 7,
587 Uo 7 (with a sotharngha). 590 Uo 8, 637 o 8. 648 o 7 (after a
sothaingha). 654 Uo 8-9 {sadaiildal Mogata sa[c^'i]), 656 Uo 5
^ trarnghuy see Burrow, BSOSy vii, 509 f.
2 Literally “to-morrow”, Skt. pmtak : pratn { — praia 361, 499), as punuk :
piinu {puna)y or puraiah : puradu ; -w represents also vSkt. -am in likhidu, BSOSy vi,
455 ; cf. dhitu, pitu, matu.
3 achid means “ take away ” as to be seen clearly from 719 o 2 ; but here, where
in line U o 3 chimnamti is used, perhaps the meaning “not care” is appropriate.
acchindati in Pali has not this meaning, but see P.H'., s.v. 5.
^ ianUy tanuvaga, tanuvaga being adjectives, see Noble, BSOSy vi, 450 f.
^ Though it is uncertain to conclude something from the order of words about the
relative rank of the karsenaia and sataiita (contrast 482 Uo 2, 4 with 590 Uo 8),
in the year 17 of king Jitugha Amguvaka, the mentioned Ricikga is a karsenavay
and in the year 24 of the same king a sadaiita (590 Uo 8. 715 o 4-5), if the two name-
sakes are identical. Is karsenaia = Skt. karsanapa ? But the stem is krisi-y though
511 o 5 shows maldprakarsena against malapraknrsana in line 4. The control of
tilling and the A-i/mc-institution reminds the reader of the rules of bestowing land to
tax-payers in the first chapter of the second book in Kaut. Arthas {Janapadaniiesa).
770
O. STEIN
(sadavida Ya[p]guasa putra Sa . . . gasa sacKi). That sadavita has
nothing to do with a denominative of sabda, corresponding to a Skt.
sabdita in the sense of “ called can be shown by the many passages
in which nam-a is used, e.g. 11 Uo 2. 28 o 2. 39 Uo 2. 49 To 3 (sach)
Sugi nama Pranga nanui). 415 Uo 6. 418 o 5. 422 o 2, 3 (five times :
in the same inscription occurs satavida in line 9, see above p. 769).
590 Uo 1, 2 (line 8 : sadavida). 719 o 2.
It seems logical to explain dusavida in an analogous sense (41 o 1-3.
r 1. 65 oA 2. C 1, 2. 66 oA 1-3. 80 rA 1. 93 o below B. 105 oA 5. E 3.
115 oD 6 under GH. K 6. 118 oD 2). That there can be no difference
in the meaning between dakivida and dasavita shows 204 where o 3
the latter, r D 2 the former word is found ; this and the sense of the
word, denoting some official, or at least somebody in charge of a duty,
becomes clear by 477 ; —
o 1 samvalsare 4 4 1 ma.se 10 1 tiva[§e 3] puratha cojhbo Lpipe. -jetha
2 yapgu e.sa amna .samg/ialidavo
3 kantamna MocKasena.sa dadavo
r 1 dasavite Jivadeyn Kalyanadhamasn ca amna [mililma 2 khi 10 2
2 dasavite Yo[n>i V]va.seria.m ca milima 1 khi 2
3 dasavite RHtrapaJa{kasa) Rutrasenasa ca amna milima 3
4 jetha yapgu
“ In the year 9, month 1 1, on the 3rd day from the cojhbo Lpipe[ya ?]
to the jetha yapgu : the corn here has to be gathered and to be given
to the Sramana Iloksasena. (r) The da sn vitas Jivademi and Kalyana-
dhaiiia (have gathered) corn (to the amount of) milima 2, khi 12 ,
the d. Yonu and Uva.sena (Upasena) corn milima 2, khi 2 ; the d.
Rutrapala and Eutrasena corn milima 3. jetha yapgu.” ^
The cojhbo give.s his order to the jetha yapgu to collect corn for a
monk, the yapgu, as can be suggested, passes the order on to his
subordinates, the dasuvitas. if they are immediately under him ; “
twice is followed by two personal names. Six da.santas occur m
the report of their collection and its amount is endorsed by the yapgu s
signature. For the pre.seut (juestion it i.s obvious that the plural
dasavite is in accordance with two individuals in each line. It is the
same grammatical situation as in 580 Uo 5 and 6, where sadavide
^ Also a cojhbo is called liJO r 1.
^ One would expect at least the satavita ; the rank and file may be ascertained
from 470, where the great king gives his orders to the cojhbo and tornga^ then it is
passed on to the yapgu. The dasavita has to do with collection and registration of
cattle and corn, etc. (cf. 148).
XUJrEEALS IN THE NIYA INSCRIPTIONS
771
104 0 1, 2 with the closing formula : ede jatnna milima 4 1 khi 10 4 1 .
apparently the quantity of corn which has been gathered from the
people of these six dasavitas. Two of them acknowledge the quantity
of corn, from eighteen men collected (108) ; 129 two dasavitas appear
beneath a group of men, under columns A and B the vestiges give :
dasalviyia bkita jamna 10 4 1 , the d. Bhita. has fifteenmen. his colleague
Samghadhana also fifteen. Kungeya only eleven. The title is found
146 o A 2. 3. B 1. 3«, 4a. rA 1. 2. 148 r. 191 o 2a ; 204 o 3 a presi of
a dasaviia is mentioned (cf. rD 2). In 205 eleven proper names are
to be read, beneath : dasarila Tamjaka 10. beneath (jam. explained
in the Index as an abbreviation for (/amnana. the number of people
liere comes to eleven, by subtracting a Tanijaka in A 3 to 10. therefore
the number 10 is given apparently with added gam. A similar case
is found 237 in which document altogether the names of twelve persons
are given, then follows "to left of columns' (n. 10) : dasavita
Malpigeya yima^ jamna (= ime jandh) 10: again 609 shows eleven
names, the last line runs ; dusavida Jimraclii jamna 10 2. so that the
dasavida Jivaraksin has included himself in the sum of 12 ; one name-
sake of him is mentioned C 2, subtracting these two, the actual number
would be 10 ; on r stands : kimna atra na esati trisa prahnre 12 10
vito pasu 1 who does not come there (will get) 30 strokes (and give)
one piece of vita small cattle ' The same name J iraracJh is found in
610 rD 1, behind it tasavita, undoubtedly another writing for dasavita,
perhaps also the same man of 609. To return once more to 41. on
r 1 the text runs : —
Apemnasa prade'jami Aviyo dasavida ata 2 Pgitasa sadami uta 3
dasavida Socara.
It looks as if in Apemna's pradesa there were sada's and dasavida s.
Lines 1 and 2 on the Obverse show that the latter stood under the
official of a ktda (e.g. Masdhigeyasa sadami dasavida Viigeya ata 2)
and. on the other hand, also under the official of a pradeja (e.g. o 3
and above r 1, where Aviyo is a proper name). It is true, the NI
do not offer the title of the official on the head of the pradeja, that
mav be due to the circum.stance that there existed only one officer
in that unit, but many sadavida's and dasavidu's. or, the pradeja
was named after some person, the name of which is given in the
‘ That re,idm>: ataonlin<; to tho (’orrection-;, p, 289.
^ nla occurs many times with aipa. go. pn.iu ■. docs it mean breeding” in
c<jntra.-.t to the vyala uta in 341, 420 ?
772
O. STEIN
preceding genitive, well-known among the official world. It is interest-
ing to see that some of the names, connected with apradeja, appear as
proper names of tomghas, cojhbos, vuryagas, apsus like Apernna {tonigha.
72, 82, 123), Lustu (cajhbo, 327, 579. vurciiga 277), Vuru (npsn 195.
I'uryaga 569), Yukto {togha 436. cojhbo 407). These imits, pradeja,
soda, dasa, and their respective heads remind the reader of the
corresponding administrative and fiscal units in Manu, \Ti, 116 if.,
with his satesa, dasesa and similar titles ; nor is pradesa and its official,
the pradesika, unknown,^ one epigraphical instance shows desddhkata
(Skt. desadhikrta)}
Even the sata occurs in inscriptions : in the Huli Inscriptions of
Vikramaditya ffi, Saka 1019 = a.d. 1097.^ in an inscription with dates
between 1104 and 1224,^ and of Bijjala, Saka 1084 = a.d. 1162,°
in connection with boundaries of granted land. It is called “ on the east,
the sata of the waste-land at Kattiyageri (p. 189), or “on the north-
east, the sata of K., on the west the sata of Raviyani ” (p. 196, also
p. 218, Ep. Ind., xviii). It seems more probable to take sata in the
meaning of an unit of administration and revenue than in that of a
“ land-measure (Ep. Ind., xviii. Index s.v.) * ; well-known are the
units of villages ranging till ten-thousands, in South Indian inscrip-
tions." But it is not correct to say ® that “ the North Indian inscrip-
tions do not refer to the grouping of villages into ten, one hundred,
and so on ” ; at least, in the Khalimpur Plate Inscription of Dharma-
paladeva ® among other officials are mentioned ddsagramikddivisa-
yavyuvahdrimh. If that suggested interpretation of the terms dasainta
and satavita as official titles, resp. dasa and sata as their local district
for 10, resp. 100 households (or villages) is correct, then it is not
surprising when the sahasrapati of Manu, vii, 117, corresponds to the
^ Cf. Archil'. Orient. y vi, 32 f.
^ Cf. Luders’a List, No. 1200.
® Ep. Ind.y xviii, p. 156, line 52 f.
* Ibid., p. 192, line 44 ; p. 193, line 47.
5 Ibid., p. 216, line 41.
® The reference to Sir Aurcl Stein'.s Serin/lifiy i, 65 (in the Index of the A /
mtade, p. 371) for the sense “ measure of land ” is not correct ; the passage runs :
“ Thus Colonel Trotter, who visitetl Wakhan with a section of the Yarkand Mission in
1874, distinctly notes that ‘ Wakhan formerly contained three “ Sads ” or hundreds,
i.e. districts, containing a hundred houses each*.” This corresponds w'ell to the
snta of the NT.
^ Kishori Mohan Gupta, The Land Sustem in South India, Lahore, 1933, 268 f.»
275 ff.
* As Mr. Gupta does, p. 27L
* Ep. Ind., iv, p. 250, line 47. Bhandarkar’s List, ibid., xx, App., p. 223,
No. 1610. Date about a.o. 810, cf. Smith, Early History, 4th ed., p. 413 f.
XUIIEEALS IN THE NIYA INSCPAPTIONS
773
tomg{h)a whom Professor Thomas ^ explains as Tibetan stoh-dpon,
commandant of a Thousand (sto«)-district ; there existed, according
to Tibetan documents,- also a civilian officer of a Ten Thousand
district (khri-dpon).
Finally, the chronological and etymological side of the question
may be touched ; the Indian terms pradeki, dasesa, satcsa are found
in Manu, in inscriptions before and about the beginnuig of the ninth
century a.d. ; on the other hand. Professor Thomas ^ has remarked
in connection with “ the use of an Indian language for administrative
purposes and . . . the adoption of an Indian administrative system "
that ‘ ‘ we can hardly assign to the beginning of the period representefl
by the Kharosfhl documents a date earlier than the third century a.d."
For the et^unology of dasarita and sakiiita there can be no doubt
that the first part is Skt. dam-, resp. sata-. The second jiart. -vita,
appears in the passive participles perf. of causatives ; but that seems
to be out of the question here. If -vita is not to be explained as an
a-stem of Skt. rid-, it may represent Skt. -arpita. as praty-arpita
becomes pic-avid-a (cf. picavidavo) : or, dasavita {-da), satavita {-da)
could correspond to a Skt. *damprta-. *satap/ta.
That sada represents also the numeral for " hundred " {sata) is
to be seen from 133 o 2. 348 Uo 6. 415 Cr 5. 589 Cr 3. where it is
found in the phrase rar.samda (or baho sadani in 133). 677 Cr 3 as
varsasata. A puzzling pas.sage is 514 o 3 ; here sadavidavya bramamua
seems to be " one to be called a Brahmana ”. but r 2 occurs mbdha.
so that a derivation of it is hardly pos.sible. Perhaps sadavidavya is to
be taken as * sraddhdpayitavya. Vi\\isaddahitahb<i. saddha yitabba .'wlierehy
the cau.sative-suffi.x -pay- is contracted into v. as vijiiapayitavya
becomes in the NI vi{m)riavidavya from vi{ai)riaveti = vijiidpayati.
In the same text 514 o 1 occurs the plural .satuni. sadatii in 133 o 2,
but without the plural-termination 149 o 2 : the noun to which the
numeral belongs does not .show any inflection. In compounds sata
{kida) forms the first or second part of them; satayn\ 107 o 1,
vamisataya 140 Uo 1. 206 Uo U2. 305 ol. 247 o 1. 646F’ol~2
{varmsatdyu-). The same conqiound shows inflected forms, like the
locative {varyididaini. 348 Uo 6). 415 Cr 5. 677 Cr 3 {' satai/nin).
110. The additive numeral for " hundred and ten " is expressed
by da.vitura {= damttara) sata 345 Uo 10 (2.v). 11.
^ K xiii, 5:i.
“ JRAS, p. ooO : 1934, j). t>7 ; <ui tax<*> and othtT ofiiciaK luncernt'd with
adinini^traticm and tax-.sur\ o\ mg. < f , ibid.. 1934. pp. f.. *04 f,. ff.
^ AO, xiii. 49.
774
O. STEIX —
1,000. For “thousand” sahasra is used, irrespective of one
or many thousand (149 o 2. 500 o 3. 661 o 3). The plural, sahaarani
is found, apart from the compound bahuliO(lisata° , in a poem with
many Sanskrit forms 514 o 1. Like -sata- also -sahasra- appears in
the formula varsasahasrami yavajivo in the locative (419 Cr 4. 579 Cr 5.
581 Cr 4—5). In the mostly Sanskrit text 511 r 3 sahasracKa is found.
Numerals over thou.sand are expressed in satasahasra- by
multiplication, preceded by hoti {Jcorji)- “ ten milhons ”, and enhanced
by bahu-, so that the whole expression runs as bahulcollisatasahasrani
in 164 o 3. 206 L’o 4. 249 o 4 (koti). 646 L o .3-4 (mistakenly written ;
sahasrahani). 696 o 3. The adjective aprameyo (npramego) following
after that extends the good wishes to the infinite (c.g. 206 Uo 4.
646 Uo 3-4).2
B. Ordix.^lia
1. “ First ” is written in many forms of Skt. prathama ; as such
it is found 220 o 1. 436 Uo 1 {nmse prathame). 291 o 5 (locat. -e).®
With unaspirated dental (pratanie) 209 ol, with media 575 L^o 1
{pradatne), a Prakrit form padama seems to be 83 Uo 3 ; the femin.
with tenuis cerebral appears in patanii mnli 437 Uo 8, and with
media cerebral 434 Uo 4, where the locative pallama samvatsaratnaii
in the Prakrit of the NI shows the uninflected form as in prathama
divasa 376 o 3. In many inscriptions prathama is used as adverb
in the sense of “ Firstly, first of all ” (140 Uo 3. 164 o 3. 247 o 2.
390 o 4 : pratama. 399 o 2 : pratharnma) ; this form must be taken
as the accus. neutr. : the comparative prathamadaro would be Skt.
*pratharnatnram or ^tnratah. used in the sense of the positive 165 o 2
and apparently 341 o 4. In the text 511. strongly influenced by
.Sanskrit, o 3 occurs the genit. plur. prathamana (prathamandtn
srdrakdnam may be intended).^
2. From the cardinal dvi the ordinal driti = dritlya is regular
and found (with vara) 45 Uo 3. 262 Uo 2. 550 o 3 (with ya[)ta] ~
yuna ?). duvi in 431 Uo 3. 432 r 2 is used as an ordinal without termina-
tion, being a locative {samratsarammi). while 703 r 2 it is a cardinal.
In most of the NI the stem bi- < dvi^ is used. biti. irrespective of
• For th(* tfnnmi. cf. Thomas, AO, xiii, o3.
^ Cf. Thomas. AO, xii, 63 ff.
^ f'f. Thomas. AO, xiii. 60.
* IVoffSsor Ludur.s (SPA 0. 10.33. 900) lK*hrvt\s prathnnyantt i-yi-n) in oOo r 1 to
he a remarkalile elorical for jtrnthnmn ; should it not (.orrespond rathor to .Skt.
'pradhij mjan/im ?
^ Cf. Pischel, Pit. Or., § 300.
NUMERALS IN THE NIYA INSCRIPTIONS
775
gender, occurs 7 Uo 2. 56 Uo 3 (with aro = vara). 58 r 1 (with vara).
140 Uo 6. 159 r 3. 198 o 2. 297 Cr 2. 305 o 6 {var.^ami. local.). 331 Uo 7.
359 r 1. 376 o 1. 437 Cr 5 (vara). As bhiti it is found 78 oC 6. 120 o 1
(vara). 123 o 2 (aja bh. divasam hoda = adya dvitlyo divaso hJiutah).
348 Uo 4. 419 Uo 2. Cr 4. 422 o 3. 425 Cr 3 (vara). 505 o 3. 571 Cr 4
(vara). 581 Uo 2. Of the fuller form the oblique cases occur like the
genitive bitiyam (187 o 8. 709 Cr 4), blntiyam (425 Uo 5) ; or direct
from bhiti- as bhitisa in the same text 425 Cr 6. The instrumental
of hiti is bitiyena (574 Cr 3). probably from bitiya-, as the genit. plur.
hhitiyana is found 577 Co 2-3. Uo 6. For the idiomatic use of the
as}Tidetic eha b(h)iti- Iranian influence has been suggested. ^ Thereby
the latter ordinal can be inflected only, as in 187 o 8. 425 Uo 5. Cr 6.
574 Cr 3. 709 Cr 4 before vamti (c. genit. : eka bkitiyasa varnti) or
sadha (c. instr. ; eka bhitiyena sadha), not so in 348 Uo 4. 709 Uo 6.
732 Uo 5. The sense may be “ one or the second ", or " one of the
two ", as the phrase eka bhitiyana vamti 577 Uo 6 makes probable.
3. In the same way as dvi-ti, b(h)i-ti for " third " tri-ti is formed
7 Uo 2. 247 0 3 (nomin.). 376 o 2 (neutr. : karya) ; the same form
as locative appears 9 Uo 3. 305 o 6. 376 o 3. With vara 45 Uo 3.
105 oE 1. 211 r 4 triti is used absolutely " for the third time ", while
in the next line triti bhagade eka bhaga corresponds to trtlydd bhagad
ekam bhaga m.
4. caturtha is "fourth" 119 o 5. 714 o 6 (locat. : masammi).
180 o B 3.^ 182 o 3. 226 o 3 (ablat. : varxade). 329 o 5 (genit. : ma.yixya)
376 o 4 (nomin. neutr.). 377 r 4-5 (catarta). 450 o 3-4. 567 Uo 2. in
these latter three instances being the nominative in the phrase :
c. varsa huda "it is the fourth year".
5. “Fifth" is yamcama (162 Cr 2); in 329 o 5 jximcatui is a
locative as to be seen from the following sasteyammi " on the fifth
day ", which appears as patncamiyammi 663 o 4, while the word for
" day " is omitted.
6. .fodha(m)ma (with ma.^e llOoinit. ; with ma.pisya 637 o 1)
for “ si.xth " must be explained as an analogous formation from so
as the following ordinals for " seventh " and " eighth ". with softening
^ Burrow. BSOS. vii, 700, superpsts “one am»ther
" Though the '’fourth” entry really eiunes in the next line B 3, the fir^^t and
second stand in A 4 and 5. The same oi eurs in 105 oE 1 : amttu friti rnm K(imrnkaf<a
muli 4, because the same man is mentioned I) :2 and 4.
® For this reading, see Corrections, p. 290, and for the meaning of cf. Sten
Konow, CII, ii, 1, p. 152, and Burrow, BSOS^ vii, 515, 783.
776
O. STEIN
and aspiration of the tenuis dental there, corresponding to Skt.
msikd-.
7. For Skt. saptamd- the NI oiler 368 o 4 satarmm (with masasya).
8. The cerebralization in Skt. astamd- is maintained in athanm
(divasa) 144 Uo 2, as in Prakrit atthama and in the Kharosthi inscrip-
tions of N.W. India.
9. ‘'Ninth” is navamma (niasami 500 o 3. nmsasya 663 o 4).
10. As in Skt. “ tenth ” is dasam-a {varsa 550 o 2), the locative
of which is uninflected in the passage 539 o 3 : sa[ratam]mi dasarna
masamnvi ; therefore, though it is not a deciding proof, dasammi
341 o 3 may be the locative of dasa-, a word corresponding to sata-,
both being administrative and fiscal units. On the other hand,
dhasamni ^ niasammi 401 Uo 5 offers a locative, which maybe a clerical
peculiarity, but also intentionally to avoid a confusion with daiami
“ in the dasa
11. For ‘‘eleventh” occurs ekadasi {masasya) 341 o 3.
12. As di'i- become.s hi-, so to Skt. dvadaki corresponds in the
NI badasi {masasya) in 599 r 1 (cf. above Card. 12 : badasa).
15. The next word in 599 r 1 is pamcadasammi which on account
of the preceding badasi masasya can be nothing else but “ on the
fifteenth day ' ' ; the same wording is met in 368 o 4 satamma-masasya
pamcadasami (15th day of the 7th month). -dasa{m)mi must be
a locative which is found also in the KliarosthI inscriptions of N.W.
India (cf. ('ll, ii. 1. p. cxiii), being the pronominal termination of
Skt. -asmia. The correct form of dasa to be expected would be
dasa>tia{m)mi, so that dasa(tH)m,i must be e.xplained as analogy to the
ten-numeral compo.sed with the numerals below ten. ekadasi - and
badasi. both with following masasya. look like feminines, but are
probablv locatives with a debilitated termination -e> i, due to the
accent on the long quantity of the ending vowel of the first part in
the compound.
18. Accordingly in a date, of which only aiasasya is left, atha-
dasami (354 o 2) mu.st be "on the eighteenth (day)''.
^ ComparinGT (M. A. Stein. Anrif'ttt Khointi, ii, plate civ) the ferm of ia in line 1
(iif?) and of yn in line 2 {khantyawmi) there it» no doubt that the reading adopted
in the text is rorrett.
2 Cf. CII, li, I, p. 22 ; Mulfidf : putnrada^e (pp. bo, 71)). But it i'' doubtful whether
pamcdmi, p. 110, i.s a pronominal termination, or influeneed bv such one, as Sten
Konow huj)pO',es (pp, < xiii and 119); paincame i.s regular, ibid., p. 28, line 1. No doubt,
local pefuliantie-i of the dialect must be taken into consideration, as in A’/, 661, for
the change of ^ or i ^ e see Xoble, BSOS, vi, 4o3, and for the use of cases, p. 451 f.
jojiierals in the niya inscriptions
777
C. Adjectiva
The compound adjectives with -guna as the second part are found
in asjTidetic connection with dvi- (142 r 1) and tri- (283 o 2. 351 o 2.
387 o 2. 676 Uo 4) in the meaning “two-, three-fold especiallv in
the formula that the duty is to be kept two- and threefold according
to the order, even at the sacrifice of the life.^ For dvili triti ga[‘)ia]
in550o3cf. above, p. 774. dvaga "double" occurs 100 r 5, perhaps
traya (or treya 1) “ triple r 3.
D. Adverbia
Iterativa. These are formed by -I’um which is found with cardinalia
as well as with ordinalia in disjunctive sense : dvivara trevara (72 r init.)
or dvi trevara (634 o 2) ; tvi trevara (719 o 2). panica sovara literally
“ five or six times but the context requires " for the fifth and sixth
time {evani ca lihhami pasava Jcaragina p. s. leklta vimrjeiiti " and
thus I vTite with regard to the small cattle, for the fifth or sixth
time I am sending a note "). For ordinalia with vara may be (juoted :
dviti triti vara “ for the second and third time " (45 Uo 3), biti vara
(56 Uo 3. 58 0 1. 437 Cr 5), bhiti vara (120 ol. 425 Cr 3. 571 Cr 4),
dviti vara (262 Uo 2), triti vara (436 Uo 5) ; also hahu vara " many
times occurs 358 o 8. varaya in eJcav. blntya (371 r 3) “ " once
more " is Skt. ekavdrakani; ekavara 534 oB 4 must probably be
separated from eknvaraya and dvivara or prathame va[re] ® in
291 o 3, 5. The former inscription contains an inventory (534 oA If.) :
[yo ■)}ia]]ii thavitaga huati (oA 1) which may be " what is to be set
up for me ", this introductory lino is followed by a list of objects
with signs of numerals, e.g. gamoca^ 3. paeliama^ I . dhamna kada
^ Thomas, A(\ xii. 44 f., n. 7 ; Burrow, JJSOSy vii, r>12. anada may bo an
equivalent for njndjutiDn, formally it is n jt'mptarn, cf. nn ; but it could be explained
also as ajhn-idh. ef. ni>a 39 Uo 4. 4l>2 Uo 3.
- See Burrow, JjSOS, vii, 7H7. J:o pi taragn is, aeeording to Burrow, be.,
pitnraga “ fat ", cf. Thomas, AO, xiii. 01. In 067 r r> raraya is not clear, and seems
to have the same sense as larayam in 206 Cr 7, where it may correspond to Skt. raraya,
as the next sentence )iia ituihi tnsa kanpyntu makes hkcly.
^ See Burrow, be.; Thomas, ^40, xiii. GO.
^ Should it be yumn{m)ra = ItDnhhnh i
' It would correspond to Skt. * pnl'-aum- which s»^ms to he unknown. In modern
Ka^t Turkestan exists the wonl hodzuma "knot, bundle”, according to Radloff^
Vfr.'-tich Wortfrhutih'i dfr Tink-1 indtkir, qazuy hnyzama. “ 1, Kleiner Teppich.
2. Gepa'-kstuck aus Wolle," lagat. httydzatna "Paiktuch ”, Menges, SPAWy 1933,
1272, s.v. Perhaps connected with that i>p<5/v/i« long nightgown-like garment
worn hy Ka.dunlrw, when ma»le of cotton cloth is> called ", SirG. A. Grierson.
PirfiiDitny nf (hr KdAimiri Lmujunip y j>. H96.
V«iL \7II. PARTS 2 AND 3. .7(1
778
O. STEIN’ —
(bow and arrows), bamboos, knives,^ silk,^ carpets,® felt rugs,^
ropes,® brass cups ® ; in B 4 appear pothi eJcavara 4 4. In 17 Uo 2-3
cam-a pothi is found which may be “ skin-garments thus pothi
ehavara would mean “garments made of one piece of cloth”.®
This cannot be applied to eka vara and did vara in 291 o 3, 5;
the meaning “ part ” is not likely on account of line 4 {did hhaqa —
eka bhaga). The context shows some resemblance to 272 o 6 ; in
both texts Kuvana and yatma Parkuiena (in 291 : Porkota, apparently
the same official’s proper name) are mentioned. Further, the reading
eka varaya is impossible, because it would not be in accordance with
dui vara, then ya ( = yd) corresponds to sd arnna and relates to amna
in line 3 (cf. yo amna . . . sa amna in 100 r 6 and o 1, 2, being the
singular), and denotes the plural, indicated by the passive anisyamti
(active : anitamti). The inscription seems to bring the opinion,
i.e. decision of the great king {cimditi [thus to read, see Correct., p. 290]
hiiati) ekavdram = prathamam vdram “ once ” in the sense of “ firstly ",
about the corn “ which will be brought into Kuvana ”,® then dui vara
= dviti vara = Skt. dvitlyam vdram “ second time ” in the sense of
“secondly” about the gathering of fifteen camel-loads of masu in
Pisaliya.^®
E. Various
For distributive expressions no adverbial formations are used,
but the reiterated cardinal as amredita-compounds : ekameka (above
p. 763) ; tre tre {milinm) 291 o 4.
Pronominal adjectives are : ekadara (307 o 7) ; Skt. pratyeka
* krataga may be .Skt. kriayab, a weapon like knife or dagger.
- kojava = Skt. kauieya, Thomas, AO, xii, 54.
® tavastagn, Burrovr, BSOS, vii, 512.
^ M. A. .Stein, Ancient Khntan, 367 ; Burrow, l.c , 510.
® rajn.
® kansiya. in note 4, -iiki-.
’’ Thomas, AO, xiii, 78 ; for poiri, cf. J. J. Meyer, ArthaS. trs., p. 656, note 1
(Arthas., xiv. 3, 178).
® pota according to Halayudha Abhidh., ii, 393 (not pota, as PW, s.v. 3, says,
only Jledini has that form and Aufrecht’s Glossary, p. 276, in his ed. of Halay.) is
“ cloth ” : varaka again is explained (PIV, .s.v. 1) as pofdrchddana “ cloak, stuff, tent ”.
“ Cf. Thomas, AO, xii, 61, 5, but here it looks like a place-name as in 272 o 6, 7.
the locative of which is regular Kuianiyammi against the locative of the term khu'van-
e.mr.i (489 r 2).
“ For the use of the cardinal with rara instead the ordinal see above ; that the
enumeration of items in a (royal ?) document was used can be seen from the fragmentary
text 376 where is found hiti, triti, calkurtha karya (o 1, 2, 4).
NUMERALS IN THE NIYA INSCRIPTIONS
779
appears as pajeka 349 o 8 and as paMeka 431 Uo 1. 432 Uo 2 (-ekam).
adha in 589 Cr 4, 5 and adha in 169 oE 3 represent Skt. ardha-.
Fractions are expressed by hhaga or bhapa (154 oA 3. 211 r5:
triti bhagade eka bhaga. 291 o 4 : dui bhaga — eka bhaga. Cf. sanmbhaga
528 Cr 1 ; sarvabhaga 18 o 2) and hy pal a (pdta) or pada, corresponding
to Skt. pdda “ fourth part 21 Uo 2 : dui pata . . . [d«] . . . ;
71 Cr 1 : dui pada . . . dui pada ; 676 Uo 5 : eka pdta .. Are pdta.
^ So already Thomas, AO, xiii, 78. In 83 Cr 6 biji pata it is ‘‘ part " (hiji padah)
as probably in 152 Cr 4 ; 666 o 2-3 and 669 o 2 has been restored in the Index,
p. 368, a.v. vamtaii (-de) to : sirasd pada (= Skt. padayoh) v.
Tathagata and Tahagaya
By E. J. Thomas
T he term tntliagata has been the sul)ject of much discussion, but
the latest treatment leaves even the etymology undecided. The
conclusion reached by the Pali Text Society's Pali-EnylisJi Dictionary
is ” derivation uncertain The question of the meaning is also left
almost as vague, for what is there said is that Mrs. Rhys Davids,
quoting Lord Chalmers, suggests " he who has won through to the
truth But a more important statement in that work is that " the
context shows that the word is an epithet of an Arahant, and that
non-Buddhists were supposed to know what it meant. The compilers
of the Nikayas must therefore have considered the expression to be
pre-Buddhistic ”. To say that they considered it to be pre-Buddhistic
perhaps assumes too much. Wo can take it that the term Mas current
among the non-Buddhist contemporaries of the compilers of the
Xikayas. but hoM' far it may have been pre-Buddhistic depends upon
M'hat non-Buddhist sect or sects used the term. Who M'ere they ?
Xaturalh' one thinks of the Jains, but investigation in this direction
appears to have been hitherto checked by the supposition that the
closest corrc.sponding Jain term is tattha-gaya. Oldenbcrg and Rhys
Davids in Vinaya Texts, i, 82, even suggested this Mord as the origin
of the Buddhist epithet. ‘‘ Considering the close relation in which
most of the dogmatical terms of the Jainas stand to those of the
Bauddhas. it is difficult to believe that tathagata and tatthagaya
should not originally have conveyed very similar ideas. We think
that in the long May from the original Magadhi to the Pali and
Sanskrit the term tatthagata or tatthagata (tatra + agata), ‘ he m Iio
has arrived there, i.c. at emancipation,' may very easily have under-
gone the change into tathagata. mIucIi Mould have made it
unintelligible. M'ere m'c not able to compare its unaltered form as
preserved by the Jainas.”
The reference given for the M'ord is Bhadrabahu's Kalpamtra
(Jinac. lb), M'here it occurs in .some limping verses at the end of a long
eulogy uttered by Sakka in praise of the tirthakaras and IMahavIra
himself. It is also said to occur in the Rajaprasnl-sfdrn, but there
as a repetition in the same eulogy uttered by another god : Va)ndanii
782
E. J. THOMAS —
nam bhagavamtam tatthagayam ihagae ; pasau me hhagavam tatthagae
ihagayam. The eulogizer who has arrived here ", i.e. has been born
in this world, is contrasting himself with the Lord who has “ arrived
there ”, i.e. who has reached emancipation. The term tatthagaya is
here scarcely an epithet. It is a very loose compound and a natural
description of the state of the Lord, just as ikagaya describes the state
of his wor.shipper. Such an isolated phrase hardly appears sufficient
to explain the origin of the Buddhist term. There is no evidence for
tatthdgata, the word forming the supposed link. Why should the
Buddhists have corrupted and made unintelligible a perfectly plain
word, and one which harmonized with their theory ? Why does no
trace of its meaning appear in the scriptural interpretations of
tathagata, if the corruption took place in the long way between the
MagadhI and the Sanskrit, that is to say, after it had come to form
a part of the Scriptures ?
It has been necessary to draw attention to this passage, because
the way in which it was adduced in Vinaya Texts has led investigators
to assume that this was all that the Jain texts had to tell about the
matter. Lord Chalmers in his important article Tathagata,'^ and R. 0.
Frankc in the appendix to his translation of the DlgJia-niJcdya, when
dealing with Jain influence, do not go beyond the evidence given by
Oldenberg and Rhys Davids. Franke calls tatthagaya the corre-
sponding Jain word, yet it was long ago pointed out by Jacobi ^ that in
its Prakrit form tathagata is common to Buddhism and Jainism, like
jina, arhnt, mahdvlra, sagnta, haddha, and even (though he does not
mention it) mahdydna? Wherever the word originated, it is evidently
necessary to find out what there is in common both in use and meaning
between Pali tathagata and Prakrit tahdgaya. The following pa.ssages
show the term as a regular epithet of the Jain leaders.
Kan kaydi aa’dhdcl appajjamti tahdgaya,
Tahdgayd. ajgmdiand cakkhd logass' analtard.
Sdyagada, i, 15, 20.
■■ How at any time can the wise ones, the tathagatas, be born
again, the tathagatas. who are free from undertakings, eyes of the
world, supreme ? "
‘ JRAS., 1898, p. 103.
2 Jaina iSulras, i, xix, xx.
® Janti vxra mahajnnam. Ayar, i, 3, 4.
TATHAGATA AJO) TAHAGAYA
783
Aho ya rdo a samiitthiehim
Tahdgaehim pmlilabdha dhammarn.
Ibid., i, 13, 2.
“ Having received the Dhamma from the tathagatas, who
exert themselves day and night.’"
Bhdsanti ege iha mdnavd u : jam ass’ alyam, lam dyantissam :
nddyam add/iatn na ya dgamissani addham niyacchanti tahdgayd u.
Aydramga, i, 3, 3.
” Some men here say, what was his past, that is his future :
not past time nor future do the tathagatas consider."
Except that the term here appears as a regular epithet of the Jain
leaders, there is little to indicate its intrinsic meaning ; but there is
no doubt that a prominent tenet of both sects was the tathagata's
omniscience, and the commentary in the Abhidhdnardjendra on the
last of the above passages has a special interest : falhdgaldni yathd-
rasthitdni tathaimvitathain jdnanti tui vibhahgujddnina iva viparltam
pasyanti ; that is, that the tathagatas know things as they have really
happened and not falsely (mtatha), just as in the Mahdparinibbdna-
sutta (D. ii, 73) it is said of the Buddhist tathagatas that they speak
nothing false, na hi tathdgatd citathani bhananti.
This is in fact the meaning which has been suggested by
Lord Chalmers for the Buddhist term, " one who has come at the
real truth,"’ but it is not the literal and primitiv'e meaning of tathdgata.
We find this meaning. thus gone," or *' having come to such a state "
not only in the Epics but also in sxich different works as Kalidasa's
Mdlankdgnimitra and the Rgvedn-prdtisdkhya. For the Epics the
evidence has been collected by E. Washburn Hopkins, who has also
attempted to define the Buddhist meaning from the epic examples
alone. 1
That the term started with the literal meaning in Buddhism could
only be assumed if the Buddhists invented it. But it is much more
likely that like arhat and such words it was taken over from the Jains
as an already established term. The meaning it would then have would
be the dogmatic sense of an epithet already applied to the Master.
There are similar instances of terms borrowed from the Jains, as shown
by S. Levi, where it is quite futile to explain the word from its
etymology, when it was borrowed as a technical term along with its
acquired sense. ^ With this agrees the fact that although there are
^ “ Buddha as Tathagata," AJP., 1911. p. 205.
“ “ Observations sur une langue precanonique du Bouddhisme,” JA., 1912, p. 495.
784
E. J. THOMAS
several passages in the Scriptures which show what it meant there,
there is no attempt to interpret it according to its hteral analysis.
Such analysis is found first in the commentaries of Buddhaghosa
and Dhammapala.i but they are unable to quote anv scriptural passage
for their interpretation, as they do for the other meanings. They
probably did not invent it. as they give some verses in which it is
implied, but they do not attribute these verses even to the poranas.-
Nor do they lead us to think that they ascribed anv special authority
to the meaning “ gone like pre\'ious Buddhas Thev give half a
dozen etymologies, and impartially take every possible or impossible
combination ; tatha + agata, tatha -h gata, tathd + OMita (he who
has not gone to sanisara), talhd + dgada [dgndanam — vacanam). tathd
+ agada (medicine). The two last, though as fanciful as the rest, have
the interest that they probably point to the existence of a Prakrit
form tahagada, which actually occurs in the Prakrit of the
Mdlavikdgnimitra.
The really significant items of the commentators are those adopted
from the Scriptures. It is not fair merely to put them aside as fanciful,
for they at least tell us what was held to be the meaning some
centuries earlier. In dealing with the Scripture passages the com-
mentators use the word tatham, the meaning of which they make more
precise by adding ai'itatham, a»annathan>. The tathdgata is, then, one
who has arriv'cd at the truth, tuthdya dgato, tatham gato, or at the
(four) truths, tathdni {tathadhamm-e) dgato. or he is one who speaks
the Truths. The passage to which Buddhaghosa refers is in the
Pdsddihi-vitta (D. iii, l.'ld). It does not use the word tatham, but it
emphasizes his omniscience and the truth of his knowledge, and gives
five reasons why Buddha is called a tathagata. (1) Wanderers of other
schools say that the tathagata (as opposed to their own tathagatas)
professes knowledge of the past, not of the future, but thev are ignorant.
He remembers as far back as he wishes, and for the future he has the
knowledge born of enlightenment. He speaks at the right time
and speaks the truth, kdlavddJ. bhdtai'ddl ; (2) he has been enlightened
with all the knowledge that can come through the senses ; (3) every-
thing that he speaks since his enlightenment is thus and not other-
Buddhaghosa on Dtgha, i, 3. Bhammapala on Uddna and Itiv. gives both
Buddhaghosa’s comment and his own.
* There is a verse in Th., i, 490, where it is said that Gotama went by the same
way (yen’ era maggena) as previous Buddhas, but no reference to tathd.
TATHAGATA AND TAHAGAYA
785
wise, taiK eva hoti no annathu ; (4) as he speaks, so lie does,
tjathdvddi, tathdkdrl, and as he does, so he speaks ; (5) he is the Master,
the unmastered, the complete seer, the controller.
There is no trace here of any explanation as thus gone ’ or as
having come and gone like his predecessors ”, and the fact that it
is also absent from the comment in the Xiddesa (which also quotes
the Pdsddika passage) is a further indication that the meaning ” tlius
gone ” is not an ancient one. On the other hand, we do not tind the
word tatham in the sense of truth or true. There is an attempt to
find significance in tathd “thus and not otherwise”, "thus doing”,
and “ thus speaking ”, but it is fairly clear that intham did not exist
for the author of this sutta. It exists, however, elsewhere, and in the
Sutta-nipdta (1115) as an adjective, " true,'’ elam ndnam ialhnm
brdhntamssa vuslmato.
The examples of tathdgata in the Sutta-nipdta, so fixr as they are
significant, all support the Pdsddika-sutta. They all refer to his
supreme knowledge. He knows the end of birth and death (467),
he has infinite wisdom (468), he comprehends all the stayings or
stoppings of consciousness, and knows hinx who stays released (1114),
and he is the supreme eye of men (347).
Both the Pdsddika-sutta and the Sutta-nipdta show in what way
the term was understood before the close of the Canon. They do not
give any intelligible etymology, nor do they give any grounds for
thinking that thev knew one. Their first interest was not an
etymological explanation, but a dogmatical interpretation ; and what
they give is far removed from the thought that the tathiigata was he
who had come and gone like his predecessors. The interpretation that
we find agrees with that of the Jains to the extent that by both sects
the tathagata was looked upon as a being of infinite knowledge.
The disputes turn, not upon the meaning of the term, but upon the
question as to which of the rival leaders was truly a tathagata.
But even the etymological sense has not led to any agreement
amongst modern interpreters. Though starting from the literal sense
thev do not accept any one of the commentators' meanings, nor do
they agree among themselves. They translate the word literally, and
then insert some qxiite arbitrary interpretation. Childers taking the
meaning ” sentient being” (satta), which the commentators attribute
to tathdgata when it refers to a released per.son, took it to mean, " one
who goes in like manner, i.e. one who goes the way of all flesh, one w ho
is subject to death, a mortal."’ For Oldenberg it means : " Der .so
786
E. J. THOMAS—
Gegangene — namlich so wie man gehen soli,” “ der Vollendete.”
Franke gives : ” Derjenige, der dieseu Weg zuriickgelegt hat, und
pragnant, derjenige, der diesen Weg (den er lehrt, zuerst selbst)
zuriickgelegt hat. " His former interpretation was : “ zur Wahrheit
gelangt,"' which agrees with that of Lord Chalmers, “ one who has
come at the real truth.” But tatliam meaning true or truth is rare,
and is evidently a secondary formation from vitatha. The literal use of
tathn in the Pasodikasiitta rather indicates that the useful word tatham
was not yet known.
So far it has been assumed that the word in question is tathdgata,
but the actual word in the Jain Scriptures is tuhdgaya. Nor could the
word have been taOidgata in the earliest form of the Buddhist Scriptures,
for it is generally agreed that their original language was some form of
Prakrit, and that they were translated into Pali in some district of
estern India. Sir George Grierson has given reasons for holding the
north-west with the centre at Takshasila to have been the neighbour-
hood. ^ R. 0. Franke put it further south, but still in a region
far removed from the home of Buddhism and from Magadhi, which
is claimed by the Buddhists as the primeval language. When the
translation into Pali took place we do not know, but even as late as
Asoka and the Bharhut inscriptions we find no trace of the hterary
Pali. Schubring confidently declares that old Ardhamagadhi was
certainly the language in which Mahavira as well as Buddha preached.^
In any case, both the Pali of the Singhalese and still more the Sanskrit
of the Sarvastivadins show that there is a Prakrit basis for their
common texts.
The word tathdgata in this earlier dialect may have been a form like
tahdgaija, tahagagn. or tahnguda. Tathdgata is the san.skritization of
a Prakrit form, but how do we know that it was a correct sanskritiza-
tion ? The form may have been a non-Aryan word entirely unconnected
with tathd and gata, but mechanically transcribed as tathdgata. This
would e.xplain why we find no attempt to connect the word with its
obvious Pali or Sanskrit sen.se until the time of the Singhalese
et\rnologist.s.
If the word belonged to some non- Aryan language, it is unhkely
that it was Dravidian. The JIagadha district is still largely surrounded
b} non-Aryan languages, but they belong to the Munda group.®
The Home of Literarj- Pali,” Bhandarkar Commemorative Essays, pp. 117 if.
l>ie Lehre der Jainas, p. 15.
See the linguistic map m the Linguistic Survey of India, vol. i, pt. i.
TATHAGATA AND TAHAGAYA
787
Several such terms have been traced by Przyluskid and their presence
in Pali would be still more likely, for however artificial a language Pali
may be, the old texts go back to popular vernacular speech.
One such example is a group of names in the Pali, which have every
appearance of being non- Aryan. In Anguttara, v. 134, there is a list
of theras, and the names of four of them show bv the variants and
corruptions that they were unintelligible to the scribes. The same four
names also occur in a list of lay disciples who had died at Xadika
(D., ii, 92 ; S., v, 358). Why they should have been duplicated like
this is a curious problem, but this fact makes it more likely that the
list is really old and not merely invented ad hoc. They are ; kafcuda
{kakudha, kakkata. kukkiita ) ; kalimbha (katinibha, kalimnm, knlibha,
kdlimba, kdralinibha, kdlinga) ; nikata [nikata, kata, nikkha) ;
katissabha (katissaha. katimmha, kathnsasaha). Besides these the
Tibetan has karkata, evidently a .sanskritized form of kakkata.
Some of these (like katissabha. *krti)n-sabha) might be referred to
Sanskrit roots, but even then the result would not be intelligible
compounds or names like any known Aryan t\^e ; and the variants
show that the scribes did not find them intelligible. The element
kati' is evidently common to several of them. The forn\ karkata appears
to be an attempt to give the word a Sanskrit appearance, but it is one
which frequently occurs elsewhere as a place-name, usually in the
form kakada or kakudha.- We also find the suffix -limba in other
non-Aryan words. The variant kdlinga is intelligible, and no doubt
the result of the scribe trying to find something familiar, and (perhaps
for the same reason) the P.T.S. edition has put it in the text.
These are examples of words which have resisted attempts to
sanskritize them. Ta/hdgata has been sanskritized. but in its use in the
Scriptures there is no trace of the Sanskrit meaning contained in
tathd and gata. The only safe ground that we stand upon is the meaning
that the Scriptures ascribe to it, a meaning that it preserved even in the
elaborate tathagata-doctrine of Mahayana. For its real etymology
more evidence is needed.
These considerations have largely a negative character, but they
may he of use in clearing away unnecessary assumptions. It is not
now possible with Franke to call tatthagaga " das entsprechende
^ ‘‘ Emprunts anaryens en indo-aryen,'’ in Bull, ^or. Hug., vols. 24, 25, 26 ; “ Noms
de villes indiennes dans la Geographie de Ptolomee, ' ibid., vol. 27.
2 Also pakudha ; this interchange of k and p has been discussed by Cohen,
Bull. ling., vol. 28, p. 81, and by Przyluski, ibid., vol. 27, pp. 218 ff.
788
TATHAGATA AND TAHAGAYA
Jaina-wort ”, or to start with Buddhaghosa's elucidations as if they
save the scriptural use and represented the historical meaning. Nor
does the historical meaning of tathdgata or iahdgaya, as it existed in the
Scriptures centuries before Buddhaghosa, show any relation to the
meanings deduced by modern interpreters from its etymology. It
may be that further examination of the Jain Scriptures will help to
decide whether the word is of Aryan origin at all.
Bvll. S.O.S. Vol. YIII, Parts 2 and 3.
Plate \
sa sa sa sa sa sa sa sam
jr
sim XU so so so sra
sya st/a sya sya sya sya sya sya sya sya
S'
dha dha dhn dhn dha dim dha dha
3 5 t
dhi dhi dhi dhi dhi m oi re
Form'. (IF S. SY . 1)11. .\Ni> C IN KiiarosthT IJi xt'mkni'. Nil. tKi
Some words found in Central Asian Documents
By F. W. Thomas
1. Hinajhasya
"jV/TR. BURROW'S interpretation {supra, Bulletin, p. 514) of the
word hinajhasya in No. 661 of KharostM Inscriptions . .
transcribed and edited by A. M. Boyer, E. J . Rapson and Senart
involves a number of conjectural or improbable elements. Thus :
(1) It is not more than probable that the Prakrit of the document
was of a form current in Khotan. The document, found at Endere,
a place in the Shan-Shan kingdom, was, no doubt, composed in some
area which at the time was subject to the Khotan king A\ujitasirnha.
But, if that area was not Endere, it may have been any other part of
the composite Khotan kingdom.
(2) If the Prakrit was of Khotan, there is yet no ground for
assigning a Saka etjmiolog}’ to the word hinajha, unless we have reason
to suppose that at the time, or previously, the current speech of
Khotan was Saka.
(3) If the word hinajha were Saka, and if it were correctly etymolo-
gized as hina, “ array," and aza, “ leader," so that the whole should
mean “general," there would be no reason for supposing it to be
a rendering of anv foreign word.
(4) If hinajha were a rendering of a foreign word, the Greek
oTpaTTjyos, which in the Indian sphere occurs only on a coin of
Azes of c. 20(i) b.c., is too remote in place and time to have in itself
any likelihood. Since Sanskrit and Prakrit were familiar in Khotan
at the time, the word translated would be Sanskrit send-nayaha.
(5) If a Khotan king had a title meaning “ General ", it would
most probably have been conferred by the Chinese court, which was
in the habit of liestowing such titles upon rulers within its sphere of
influence.
In note 4 to the edition of the text attention is called to a possibility
of reading the last syllable of hinajhasya as rye or mje. Wliatever
790
F. W. THOMAS
may be thought of these alternatives, it is certain in any case that the
syllable is not sya. In the document the syllable sya has nine
occtirrences, and in all these it is formed by the same strokes to the
same effect. Moreover, there are of sa seven occurrences, of sam one.
of si one, of su one, of so three, of sra one ; and in all these also the
formation of the s is the same as in the sya. The here accompanying
photographic reproduction, wherein the occurrences are grouped,
will make these facts patent to every eye. The photograph shows
also the occurrences of dha and ca, the only other aksaras having any
similarity to the one in question, which is exhibited with a *.
This confrontation affords the certainty that the syllable following
kinajha is not sya and a high probability that it is dhe, since it lacks
the curve (derived from a loop) in the head of ca.
At the recent Congress of Orientalists in Kome Professor Sten
Konow, having accepted the reading proposed above and urging that
the aksara following the dhe is not a but va, a possibihty indicated in
the Editors’ note 5, propounded a redivision of the words in the
form : hinajha dheva Vijidasitnhasya, with dh for initial d in dheva
according to the practice followed in the document. This does indeed
seem to posit in dheva a serni-dialectical form, since the practice of the
document replaces Sanskrit e by i (e.g. in dhinati, 1. 7 = denati)
and dhiva, = deva, is perhaps actually to be found in Bahudhivd (1. 7) :
it also conceives the existence of a king with the name Vijitasimha,
which I should not have ventured to do. If it should prove necessary
to accept a reading dheva = deva, we might perhaps avoid the second
inconvenience by understanding °devavijitabi7yha as a compound,
which would leave the personal name Avijitasirnha intact. Hinajha-
dheva might then be a surname with deva for second member, as in
Svarna-deva (of Kuca), Vasu-deva, and the many Khotan names in
-de noted in JRAS. 1930, pp. 295-6.
In these circumstances it is satisfactory to have at least the
certainty of the equation hinajha — send-nl or send-pati, which is
furnished by the anne.xed note, kindly contributed by Dr. H. W.
Bailey : —
hiaaysa general ’
Stein MS. Ch. c. 001, lines S51-1058, contains the complete Sumukha-
dharanl = Tib. Kanjur, rgyud xiii, 416^ .seq. (Narthang edition). In turn
divine beings come forward offering long life to the reciter of the Sumukha-
dbaranl beginning with Brahmdnd gyastd, Tib. IJiahi dhan-po brgya byin.
WORDS FOUND IN CENTRAL ASIAN DOCUMENTS
791
Sixth is Mahisvard gyastasai, Tib. lhahi dban-phyug chen-po. then the
following passage : —
986. ttiya va skqndha ays^nai
mista hinaysa k^ma halai gyasta
987 baysa vye hasta amjala
dastyam aurga tsve ii gyasta
l^ysa tta hve.
Then Skandha ^ the youth ^ the
great general, where the Bhagavdn
■was, thither with hands forming
the anjali he went with reverence
and thus spoke to the Bhagavdn.
Hence hinaysa (haind d- az-, cf. Av. gav-dza-, nav-dza [read
= sde-dpon {sde = Skt. send, dpon = Skt. pati).
2. Kharosthi s < dh
Professor Liiders’ identification (Berlin Academy Sitzungsberichte,
1933, pp. 1000-1) of the form masu, in the Kharosthi documents,
with Sanskrit madhu, an identification to which a personal adherence
has been expressed in Acta Orientalia, xiii, p. 60. n. 2, encounters
a certain reluctance by reason of the unfamiliarity of the transition
from dh to s. The change had previously been remarked by
Professor Rapson (index) in the word asimatra, with variants admatra,
ajhimatra, which can scarcely be different from adhimatra. which
also occurs. Consideration of the nature of the change, which
Professor Liiders refers to Iranian influence, must depend upon further
information concerning its range. It is therefore of interest to note some
examples occurring in the Indian sphere, probably all at an earlier,
and one at a very much earlier, date. Those which I have found are
the following : —
1. sasu = sddhu (Asoka Edict I, Shahbazgarhi, 1. 2) in sasu-
mate = sddhu-mata, etc.
2. bosi-satva = bodhi-sattva (Taxila Silver Scroll, 1. 3 ; see
Kharoshtkl Inscriptions, ed. Konow, p. 77).
3. Bosa-varuma = Bodha-varmmi (Lahore Museum Halo
Inscription, ibid., p. 115).
In Brahmi inscriptions I have not found any example, and, so far,
therefore, the probability is that the phenomenon was peculiar
* Skt. Skanila “ god of war ”, called Skandhakumar- in Tocharian 370, 5, with
dh a.s here in Saka.
* Translates Skt. kumhra.
Tib. 4216, i. de-nas sde-dpon
gzonus I bcom-ldan-hdas ga-la-ba
der logsu thal-mo sbyar-ba
btude I bcom- 2 -Idan-hdas-la hdi
skad-ces gsol-to 1 1
792
F. W. THOMAS
to dialects of the north-west. By reason of the dates its origin cannot
be placed in Chinese Turkestan. Its real history, in which some
Iranian influence may have been exercised, demands for its elucidation
further material.
3. Nictri, NicHatra, Picara, Picaveti
These four words, equivalents of Sk. nak^atra, Sk. pratyarha,
Sk. pratyarpayati, have a common feature in so far as they all present
the vowel i in place of a : three of them — nicliatra, picaveti, picara
(sometimes pimcara) — have several occurrences, and the first occurs
side by side (in No. 565) with the normal nacMatra. The forms support
each other and show that in the milieu wherein they arose — people
of the Shan-shan kingdom, of Chinese Turkestan, employing for some
purposes an Indo-Arvan tongue — a tendency to pronounce a as i
was able, under some accentual conditions, to fulfil itself w’hen the
following consonant was c (cc) or ck. But something should be said
concerning three of the four words.
Siciri, which occurs in the document No. 677, may be regarded
as an adjective from the nacira of Nos. 13, 15, 156, 509 on the ground
of (1) the rarity of words containing c, (2) the prima facie appropriate-
ness of the meaning " not connected with army (send) nor connected
with hunting (nacira) " in the passage : tasya bhumasa na seni na
niciri harga anti, where, however, the meaning of harga is unascertained.
But naturally the suppo.sition i.s a conjecture.
Pic[c]pra. occurring in the complimentary phrase —
picara-divyn-var.pi-kitaya-prainnna (Nos. 107, 247, etc.)
and used also in —
lekha pruhada prenisama yo teJii picara syati (No. 288).
We will send a letter and present such as may be worthy
of you
and similarly, no doubt, in the defective pa.s.sage —
atra rilnaji]da ya ja sa Priyamac . . . picara niyati (No. 377)
can scarcely be other than = Sk. pratyarha in the phra.se (Mahdvastu. ed.
Senart I. p. 467) pratyarha-makhya “ happiness according to desert ”.
But the loss of r seen in pic{c)ara < pric(c)ara requires consideration.
The preposition prati shows irregularity in the documents, having
variant forms pnli, padi, and prati (see Professor Kapson's Index,
and note pad’ cha = pratyckn), in regard to which we may conjecture
different e.xplanations. But in the case of pic{c)ara it seems reasonable
WOKDS FOtJND IN CENTRAL ASIAN DOCUMENTS
793
to admit a dissimulative influence of the following r. A dissimulation
in the opposite direction may have worked in the case of canidri-
ka(r)mamta (No. 272) and cdmdri-l:a{r)mmamta (No. 714), discussed
in Acta Orientalia, xii, p. 46, n. 3. But independent weakness of r
preceding consonants seems to be evidenced by such forms as hit-
ersi = hit-aisi (No. 511) and Saka-KhotanI Armdtdya = Amitdyu
(Saka Studies, by Sten Konow, p. 34), etc.
In the case of the numerous forms of the verb pic(c)aveti
= pratyarpayati it seems diSicult to have recourse to dissimulation,
smce both r’s are wanting. If we nevertheless suppose a loss of the
second r, we might at least expect the p to remain, giving pic{c)apeti :
in Prakrit the forms of the verb paccappinai (Pischel, Grammatik
d. Prakrit-Sprachen, § 557) always have pp. But it seems probable
that in the dialect which was the source of the verb pic{c)aveti there
had been a confusion of arpayati (from ar) and apayati (from ap)
and that the immediate predecessor of pic{c)aveti was piccapeti
< palyapayati < patyarpayati. That confusion at some early time had
taken place between arp and dp in connection with prati appears
from the verb-stem of paccappinai (i.e. °noti), in which Jacobi {Kuhn’s
Zeitschrift, 35, p. 573, n. 2) rightly recognized a suffixal -ina, rejecting
Pischel’s supposition of a denominative from arpaiia.
As regards the meaning of pic{c)aveti it may be observed that the
sense of “ dehver ”, render ”, not “ render back ”, is regular in
Prakrit (nivedane) and is found in Sanskrit {Rayhu-vamsa, xv, 41,
ap B. and K., s.v. ar).
4. DapicI
This place, named in the Saka-KhotanI document published in
Two Medieval Documents from Tun-Huang by F. W. Thomas and
Sten Konow (p. 148), was not identified either by the editors or in the
notes pubhshed by Mr. G. L. M. Clauson in JRAS., 1931, pp. 297-309,
where I-cu, Phucamni, and Tsirikyepi, mentioned in the same con-
nection, seem to be correctly referred to Hand, Pichan, and Sirkip.
As DapicI seems to be the first stage on a route from I-cu (Hami),
it can hardly be other than the place known to the Chinese as Na-chih.
at a distance of 120, or 310, li west of Hami and at present named
Lap-chuk (in Arrowsmith's old map Labezu). As Professor Pelliot.
who visited the place, points out {Journal Asiatique, xi, vii (1916),
pp. 116-19), it was foimded, probably during the sixth century a.d.,
VOL. Vm. FARTS 2 AND 3. .51
794
WORDS FOUND IN CENTRAL ASIAN DOCUMENTS
by colonists from Nob (Charklik), who named it after their home-
city, the original form of the name being Nap-ciJc. Professor Pelhoi
discusses the phonology of the change from Nap to Lap, which i-
familiar in other cases {Lop from Nd), etc.).
It seems that we must recognize in Dapicl a third form Dap (the
i of Dapi being merely a Saka-Khotani orthographical (?) expedient,
in N aki-cMttipu, etc.). In Chinese transliterations syllables (wa, etc.)
beginning with n are frequently used to represent foreign d-, and
inversely Chinese initial n appears in Tibetan sometimes as hd (i.e.
nd, JRAS., 1926, p. 525 ; 1927, p. 305). Whether m Dapici (eighth
century a.d.) the d existed in the local pronunciation as a transition
stage between n and I or betrays a Chinese intermediary in the
information, we have no means of ascertaining.
It may be remarked en passant that, as the places named in the
passage here cited seem to form a not very widely extended group,
further identifications may be expected. Thus Ttiydki may be Toyuk,
an ancient site, as is shown by archaeological investigations (Sir A.
Stein, Inyierrnost Asia, pp. 613-15), and possibly Sakdhi may be the
Soga given in a map (Hedin-Herrmann, Southern Tibet, viii, pi. xxiii),
and others suggest themselves. But there is little profit in such
identifications, unsupported by early forms of the modern names.
Sanskrit a-kseti and Pali acchati in Modem
Indo- Aryan
By R. L. Turner
TN a notice ^ of an article ^ full of new facts and invaluable suggestion
like all those of the great scholar, to honour whom this volume is
designed, I ventured to suggest that the origin of Pali acchati was to
be looked for in Skt. a-kseti. The assumption of an Old Indo-Aryan
form with ks was necessitated by the Kashmiri chuh ‘ he is ’ : for Kish.
ch corresponds (except in loanwords) to Skt. ks ; Skt. {e)ch > Ksh.
dental affricate ch (ish). Professor J. Bloch,® while accepting the
identity of Ksh. chuh with MidlA. acch-, appears to recognize the
difficulty of equating Ksh. ch with cch, but passes over the possibility
of acch- being derived from a-kseti. A retinn, then, to this much
discussed question of the derivation of acch- is perhaps permissible.
Various forms from five Sanskrit roots had previously been
suggested as the origin of acch- : stha-, gam-, r-, as-, as- ‘ to be ’.
1. d-sthd- by A. F. Pott,* approved by 6. A. Grierson and A. F. R.
Hoernle,® who supposed a metathesis, *d-thsd-.
2. gacchati with loss of initial g- by A. Weber,® E. Muller, ’ and
E. Leumann.® This was rejected by E. W. A. Kuhn ® and R. Pischel.*®
3. rcchdti by Pischel.** This has more recently received the support
of Sir George Grierson himself.*®
4. Hemacandra *® and Pali grammarians ** referred acch- to Skt.
as-, which was accepted by R. C. Childers,*® Pischel,*® P. Steinthal,**
1 BSOS. V, p. 137 ff.
^ Garbe-Festgabe, p. 24 ff.
® L' Indo-aryen du Veda aux temps modernes, p. 53.
* Die Zigeuner in Europa und Asien, i, p. 459.
^ A Comparative Dictionary of the Bihdri Language, p. 93 ff., where appear most
of the references to discussions up to 1885.
‘ Das Saptagatakam des Hala, p. 556.
’ Beitrage zur Grammatik des Jainaprdkj-it, p. 36.
* Das Aupapdtika Sutra, p. 93.
® Beitrage zur Pali-Grammatik, p. 97.
Beitrage zur vergleichenden Sprachforschung herausgegeben von A. Kuhn, viii,
p. 144.
** Grammatik der Prakrit-Sprachen, § 480, where previous references are recorded.
Garbe-Festgabe, p. 24 ff.
” iv,215.
** Saddaniti, ii, Moggalldna-Vydkarana, v, 173, quoted in Critical Pali Dictionary,
s.v. acchati.
A Dictionary of the Pali Language, s.v.
Hemacandra' s Grammatik der Prdkritsprachen, p. 155.
*’ Specimen der ydyddhammakahd, p. 45.
796
K. L. TURNER —
and Hoernled To explain the presence of cch Pischel ^ had earlier
proposed an inchoative with IE. suffix -ske- [*es-s^e-], which was
accepted by W. Geiger.® Before that G. I. Ascoli * had proposed a
future *atsyati or *dtsyate. (This appears to be accepted by
D. Andersen and H. Smith for Pa. 2 sg. fut. acchasi ®). Two years later
V. Trenckner,® proceeding from Pa. aorist acchi suggested *dtsit,
aorist of ds-.
5. Vararuci,^ Kramadfsvara,® Ramasarman,® and Markandeya ®
referred acch- to as- ‘ to be ’. E. Kuhn/® again to explain cch, took
the inchoative with IE. -ske-. In this he has been followed by S. Levi
and A. MeiUet/^ J. Bloch/® Andersen and Smith.®* E, Senart and
Johansson/® interpreting the acchariiti (ayhcharhti according to Senart)
of Asoka Shah, v, 11, as a future, set out from a future of as-, viz.
*atsyati.
The basic meaning of acch- is ‘ to abide, sit, remain seen clearly
in Pali (cf. samacch- ‘ to sit down together ’) and Prakrit ®® and preserved
down to the modem languages in Gypsy oc(A)- and Old Hindi dchnd?*
Therefore on the score of meaning alone we may dismiss gdcchaH
and rcchdti.^^ Nor is it possible to separate acch- ‘ to abide, sit ’ from
acch- ‘to be *, used so widely in the modern languages both as the
^ Comparative Grammar of the Gaudian LanguageSy p. 366 ; or to as;
- Gottingische gelehrte Anzeigen, 1875, p. 627 f., and Beitrdge zur vergleichenden
Sprachjorschung, viii, p, 144.
^ Pali Literatur und Spraohey § 135.
* Studj Criticiy p. 352, note 49.
® Crit. Pali Diet., s.v. acekati.
® Pali Miscellany, p. 61.
^ xii, 19.
® iVy 10 .
® Quoted by Grierson, Memoirs of the As. Soc. Bengal, viii, Xo. 2, p. 88.
Quoted by Pischel, Gr. Pkt. Spr., § 480, note 6.
MSL. xviii, p, 28.
La formation de la langue maratke, p. 289.
” Crit. Pali Diet., s.v.
Les Inscriptions de Piyadasi, p. 138.
1'^ IF. iii, 210.
iSee especially the evidenc-e collected by Grierson and Hoernle in Comp. Diet.
BihdrT, p. 93.
See especially J. Sampson, Dialect of the Gypsies of \yale8y pt. iv, pp. 1-2.
Syara Sundar Das, Hindi-l^abdasdgary s.vv. achnd, dchyid.
The meanings indriynpralaya and murtibhdva given by the Dhatup. for r- and
referred to by Pischel Gr. Pkt. Spr., § 480, are presumably due to the use of r- with
abstract nouns in the accusative, e.g. ynddharangatdm r- (Xalod. 2, 10) ‘to become
the battleground of It is true, as Grierson in Garbe-Festgabey p. 24, points out, that
‘ to go ’ may thus develop into * to become as in gdcchati >• Ksh. gachun ‘ to become
But acchati means ‘ abides not ‘ becomes *.
798
R. L. TUSNER
substantive verb and as an auxiliary to form participial tenses. The
beginning of this development is already found in Pali where, as
T. Rhys-Davids and AV. Stede ^ point out, the use of the present
participle with acchati is parallel to the similar idiom in Sanskrit of
the present participle with aste. Thus it would appear that IE.
*es(s)fe- ‘to be ’ has little chance of being the origin of Pa. acchati
‘ abides, sits ’. We are left therefore with the forms of as- ‘ to sit ,
*dcchati (< *es-sJce-) and *dtsit or *atsyati.^
All these forms have Skt. cch or ts(y). The group ts(y) falls together
with cch over the whole Aliddle and Modern lA. domain except in a
small group of Dardic dialects.®
Though it may be noted that none of these three forms is actually
found in Sanskrit, semasiologically any of them would be satisfactory
and on the phonetic side would account for the forms of Pa.Pkt.
acch- ; Gypsy ac(h)- ; Garhwali, Kumaoni, Nepali ch- ; Assamese
as- ; Bengali dch- ; Oriya, Maithili ach-. Old Hindi ach- ; East
Rajasthani (Jaipurl, Harautl, Banjarl, East Malvi, NimadI, Central
Bhll dialects), Gujarati ch- ; Marathi as- ; ElhandesI s- ; KohkanI,
Halbi as-.
But, inseparable from these, there are in both the Middle and
Modern languages forms which caimot be brought under the formula
of a Skt. cch or ts{y). As already pointed out, Kashmiri has chuh ‘ is
and in Kashmiri ch is derived only from Skt. hs : both Skt. {c)ch and
Skt. ts > Ksh. ch (dental affricate).* On the other hand, in the East,
where Skt. ks > kkh, beside the ch forms of Garh., Kum., Nep., Maith.,
Beng., Ass., Or.,® forms of the substantive verb are to be found with
kh. Old Maithili had akhalu ‘ was ’, with which S. K. Chatterji,®
^ Pali Text Society's Pali-English Dictionary^ s.v. acchati.
Pott s a-stha- is impossible on phonetic grounds, and was discarded by Grierson
and Hoemle.
® E.g. 8hina distinguishes Skt. {c)ch and ts as ch and ts{h) ; ckijei < chidyate,
chal <; chagaidhf but uts ■< utsdh, hatsho <c. vatsakah.
BSOS. V, p. 138. For ts{y) we have mtocA” < vatsakaht mack -< maisyah-
In the Bihari {except Maithili), East and West Hindi areas we have no evidence
for the existence of Uch- as the substantive or auxiliary verb. It is not so found to-day.
In Old Hindi ach- was a verb of fuller meaning ‘ remain, be found, exist Syam Sundar
Has in the Hindi SahdakoSy s.w, acknd, dchndy gives references to JaisI, K.abir, and
Bihari : it thus belonged to the vocabulaiy of literary AwadhI and Braj ; and may
have entered from the East Rajasthani dialects, where to-day it provides the
^bstantive and auxiliary verb. That in Old Awadhi was dh- (see, e.g. Grierson and
Hoemle’s Index to the Ramdyan ofTuhi DdSy pp. 23, 31).
/.I of JyotirUvara-kavisekkardcurya (reprinted from the Proceedings
the Fourth Oriental Congress, vol. ii), p. 69.
SANSKRIT A-KSETI IX IIODERX IXDO-ARYAX 799
although he leaves it unexplained, rightly compares Bhojpuri kfie
' is naikhe ‘ is not ’A To these may be added Nagpuria nakhi ‘ am
not Madhesi naikhi.^
Much farther to the East, the existence of a kh verb in Bengali
is attested by otherwise inexplicable forms of the Cakma dialect of the
Chittagong Hill Tracts. Here the substantive verb is ; —
Present Past
Sg. and pi. 1 dgi I am Sg. eluh I was pi. elan
2 dgas ele eld
3 dge el eldk *
This dialect generally has unaspirated voiced stops corresponding
to intervocalic voiceless stops, aspirated or unaspirated, of Bengali.
The following examples occur in the specimen ® : ghadaki = ghatahi,
egcUtar = ekatra, cdgar = cdkar, nigili — nikal-, bhidar = bhitar, afiudi
= anuthi, hdd{-at) = hath. Therefore dgi may be derived from *dkhi,
and el < *dgil < *dkhil.
Many years earlier J. Beames * had perceived the unity of the kh
and {d)ch- forms. He invented a Skt. root aks- ‘ to appear ’ (which he
based on dksi ‘eye’) to account for the correspondence kh = ch.
Johansson ’ rightly rejected the non-existent aks-^ but with it wrongly
threw over the identity of the kh and ch forms of the verb.
Since Pa.Pkt. acch-, Ksh. chuh, OMaith. akh- can only be umted
under a common form containing Skt. ks, it is imperative to examine
again the Asokan passage in which E. Senart ® read aihchamti, G. Buhler
and A. C. Woolner ® achamti ‘ are ’ or ‘ will be ’, Shah., v, 11, maa
putra ca nataro ca jKiram ca tena ye me apaca achamti avakapam my
sons and grandsons and after that those who are my descendants to
the end of time '.
In the other four versions in which this sentence occurs there is no
verb expressed :
Man. 7 naa putra ca natare ca para ca tena ye apatiye me avakajpam.
^ Grierson, Linguistic JSurvey of India, v, 2, p. 51, and Seven Grammars of the
Bihdri Language, p. 41, gives only the negative na{h)ikh-.
* LSI. V, 2, p. 280.
’ LSI. V, 2, p. 305.
‘ LSI. V, 1, p. 324.
‘ Ib., p. 327-^9.
® Comparative Grammar of the Modern Aryan Languages of India, iii, p. 183 (cf.
1, p. 192 f.).
' IF. iii, p. 209.
* Les Inscriptions de Piyadasi, p. 138.
* Aioka Text and Glossary, p. 54.
800
R. L. TURNER —
Gir. mama puta ca pot a ca param ca tena ye me apacam dvasam-
vatakapd.
Kal. mama puta, cd natdle ca palam cd tehi ye apatiye yne dvakapam.
Dhau. ye me putd va nati va . . . m ca tena ye apatiye me dvakapam.
This renders it highly probable that the verb in the Shah, version
is the substantive verb, and at the same time that it is the present
tense rather than the future {achamii < *aisyanti) which Senart and
Johansson ^ maintained it to be.
But E. Hultzsch,^ following A. M. Boyer,® notes that the second
aksara is that which corresponds to Skt. ks and he transliterates as
Es, though without prejudice as to its pronunciation. E. J. Eapson *
agrees with Sten Konow ® in interpreting the corresponding form with
a line over it of the Kharosthi documents from Niya as a compound
aksara, viz. ks. Hultzsch, now reading aksamti and finding no Sanskrit
equivalent for this word, was apparently tempted to read the first
aksara, which presents certain difBculties, as vra, making vraksarhti.
This he interpreted as future of vraj-, *vraksyati (present stem in
Shah. vrac{c)-). Neither this form nor this use is elsewhere attested for
Skt. vraj-, Pa. vaj-, Pkt. vaj-, vajj-, vacc-, or for the modern languages.*
A close examination of the plate given by Hultzsch seems to show that
the reading supported by Senart, Buhler, and Woolner as a is correct.
All the examples of v have a rectilinear angle made by the horizontal
and perpendicular lines forming it. This one has the typical curve or
hook of the aksara for a. Only at the bottom appears to be a stroke
which Hultzsch read as r ; but this is possibly a meaningless mark
on the rock. The word then is aksariiti ‘ exist, are ’, and in it we have
a word which corresponds exactly with a form of the verb ‘to be ’
found in the Kharosthi docmnents from Chinese Turkestan, namely
aeJi-, which as we have seen is probably to be read rather as aks-.
Of the three certain examples ’ two are the auxiliary and one the
substantive verb ; —
^ IF, iii, p. 210.
* Cll.f Tol. i (new ed.). Inscriptions of Aioka, p. 55, note 5.
* JA. 1911, p. 422 f.
* Kharosthi Inscriptions discovered by Sir Aurel Stein in Chinese Turkestany p. 302.
® Deutsche Literatur-Zeitung, 1924, p. 1902 ; CII. vol. ii, pt. i ; Kharosthi
Inscriptions, p. cx.
® See Comparative and Etymological Dictionary of the Nepali Language, s.v.
bagnu.
’ E. J. Rapson and P. S. Noble, Khar. Inscr. Turkestan, pt. iii. Index, s.w. achaiit
achati, achatu. Of these the first, in No. 506, read by the editors acKati (not acHati)
according to a communication from Mr. T. Burrow, to be read as aja vt,
which the editors give as an alternative.
SANSKRIT A-E.vETI IN MODERN INDO-ARYAN 801
No. 83. yarn kola tuo nigata rayadvarammi u[kasidavo] acHati
‘ when you must go out down to the king's court ’. yati tuo na
ukasida\yo'\ acKatu ‘ if you should not have to go out ’.
No. 188. . . . [yo] . . . tahi karya acl'iati, sacJlami \a\ham tahi
karya karamnae ‘ I shall be able to do what business there is of yours '.
Contaminated with hoti and huad ( < bhdvati) this verb appears as
hacK-^ in seventeen documents. Like acK- its use in conditional and
relative sentences of the type yadi bkudartka eva JiacKati ‘ if the fact
is so corresponds closely with the one instance from Asoka jKirarn
ca term ye rne apaca aksarhti.^ As a form of the modus irrealis, it is used
m the same way as siyati, which may account for its appearance in
one document (No. 4), like that of acliatu above, as hachatu ; yadi
Uhl na visarjidae hachatu ‘ if the camel should not be sent ’.® On the
other hand bhavisyati sometimes replaces it, and for the same reason
as led Johansson to describe Shah, aksarhti as a future, hachati could
be conceived of as a future *, whence doubtless the learned spellings
in Nos. 223, 366, and 578 as hachyati. Not only its use, but also its
form would give it the appearance of a future of the type Skt. vaksydti,
bhaksyati, iaksyati (cf. sacKami above), etc. In the same way in
Prakrit the present stem gacchai became a future,® through the influence
of futures like lacchal, bhecchai, mocchai derhnng from Sanskrit forms
with -psy-, -tsy-, -ksy-.^
In Prakrit a similar contamination of ho- with the descendant of
dkseti attests the existence of a kkh form of this verb. The reality of
AMg. hokkhal which, though frequent, Pischel ’ wrongly sets down as
a false reading of a form resting on a *bhosyati, is proved by the
^ Rapson and Noble, Khar. Inscr. Turkestan, pt. iii, Index, s.w. hachati, hachatu,
hachyaii. T. Burrow, JRAS. 1935, p. 669, considers aeH- to be derived from hach-
with the loss of initial h- seen occasionally elsewhere in these documents, e.g. astammi
= ha°, uhati = huati. But there seems no reason to doubt that ach~ may be the
original form.
^ See above, p. 799.
® Cf. the invasion of the 2 sg. optative by the imperative ending -su, Pischel,
Gr. Pkt. Spr., § 461. But according to Burrow in his thesis ^4 Grammar of the Language
of the Kharoathi Inscriptions (deposited in the Cambridge University Library) these,
with other forms in -tu. are 2 {-tu < tuvdm).
^ F. \V. Thomas, Acta xiii, pp. 61-2, translates two examples in No. 165
as futures ; yo puna tahi karyani hachamti ‘ whatever requirements of yours shall
come ’ ; yo atra subhasubha-ia prarrti hachati ‘ whatever occurrences of good and bad
there shall be
® R. Pischel, Gr. Pkt. Spr., § 523 : gaccharh, gacchimi, etc., though Pischel’s proposed
*gaksydmi has no foundation.
® BSOS. vi, p. 535 ; cf. Asoka KalsI, etc., ka{c)chati replacing ambiguous *kassati.
^ Gr. Pkt, Spr., § 521.
802
R. L. TURNER
existence of the ' strong auxiliary verb ’ holch- ‘ to be, become ’ in
Bhojpuri ^ and in the Magahi west of Gaya.^ It might, it is true, be
urged that a future hokkhai was formed direct from the root ho- after
the type bhojate : bhoksyati (cf. Pkt. hhoyawa- : bhokkhai = hoyavva- :
hokkhal). But Magahi of South Patna and Gaya itself has emphatic
forms of the present of the verb ‘ to be ’ which correspond exactly in
form with hachaii of the Kiharosthl documents, viz. 1 sg. haki, 3 sg.
hakai, 3 pi. hakhin : to this last form Grierson ® adds the significant
note : ‘ Forms such as this, containing kh, are much used by Kayasth
women ’. For women especially preserve archaic forms.
A similar, but probably independent, contamination of the two
stems exists in the ' optative ’ of the verb ‘ to be ’ in the Nuri dialect
of Asiatic gypsy : —
sg. 1 hdcam pi. hocan
2 hud, hosi hoces
3 hucer hocdnd *
In the Middle Indo-Aryan, then, of Shahbazgarhi and of the
Kharosthr Documents we have words for the verb ‘ to be ’ — aks-
and dch- — which correspond in form and meaning with Ksh. chuh
and Maith. akh-, and attest once again that the cch of Pa. acchati is
derived from Skt. ks, as Beames saw. If he had not had recourse
to the imaginary root aks-, he would have found its obvious origin
in Skt. d'kseti ‘ stays, remains, exists
Even in causative stems MidlA. e < Skt. aya was often in Prakrit
replaced by a.® In most modern languages the inflection is identical
' G. A. Grierson, Seven Gr. Bihari, pt. ii, p. 51.
‘ Ib., pt. iii, p. 31.
’ Ib., pt. iii, p. 31.
‘ E. A. S. JIacalister, The language of the Xavar, p. 36. J. Bloch, who was the first
to recognize the existence of acvh- in this dialect, Journ. Gypsy Lore Soc., 3rd ser., xi,
p. 32, explains this paradigm a little differently as a compound tense containing le
radical (on I’absolutif ?) du verbe ho-, suivi d’un verbe conjugue signifiant lui-meme
■ etre
^ a-icsi- according to Gra.s.smann occurs six times in RV. ; BK. give four references
- und none to any subsequent text, aksit- once in RV., dndksit- in jSBr. An
*dksaya- m. ‘ resting-place ’ possibly survives in Hindhi dkhero m. ‘ bird’s nest
1 have found no other surviving verbal form of ksi- in the modem languages, ksema-
(Pa.Pkt. khema-, Khar. Doc. cKema-, Si. kht f. ‘ welfare ’, Guj. khem n. (?) ‘ well-
being , ilar. khev m. ' evil accident ’, Sgh. semin, hemin ‘ slowly, softly ’, kema ‘ magic
to avert mischief ; yogaksemd-, see Xep, Diet. s.v. jokhim) and especially ksetra-
(see A ep. Diet. s.v. khet) have had a considerable fortune.
‘ Pischel, Gr. Pkt. Spr., § 553.
SANSKRIT I-KSETJ IN MODERN INDO-ARYAN
803
for both -a- and -aya steins d Already in Asoka in a verb without
causative meaning Kal. has pi. kalayiui beside sg. Icaleti ^ (cf. Shah,
pres. part, karamtam ® beside karoti). In Pkt. nei ( < ndyati) stands
in contrast to dnai ( < dnayati). Since the suflhx -e- ( < -aya-) is
predominantly associated with transitive (causative) verbs, a pre-
eminently intransitive verb hke dkseti would aU the more easily be
replaced by dksati, a process doubtless assisted by analogies within
the form-group itself, such as carita- : carat i = aksita- : dksati.
Nevertheless, in the conservative languages of the West and
North-West, namely SindhI and Lahnda, there is evidence that this
verb contained an e. Sindhi distinguishes the conjugation of -a- and
-aya- verbs of Sanskrit in the old present ; —
sg. 1 card (cf. Pkt. cardmi)
2 car-e, -i
3 care
pi. 1 card
2 caro
3 cardn'
cdrid (cf. Pkt. cdremi)
cdr-ie, -e, -I
care
cdrid
cdrio
carird
Although the second or -i- conjugation is otherwise confined to
transitive verbs, yet the substantive and auxihary verb dh-> which is
probably < dkh- *, belongs to this conjugation : —
In the Laru ® :
sg.
1 dhid
pi. dhid
2 dJi-e, -i
dhio
3 dke
dhird
sg.
1 ayd
pi. dyd
2 at
dyo
3 dhe
dhin'
This, so contrary to the conjugational system of Sindhi, can only
be the result of a MidlA. form with -e-, such as might rest upon a Skt.
dkseti. The retention of -e- in the iVDdlA. ancestor of Sindhi in
contrast to Pa. acchati is paralleled by dinid, etc. (<Cdnemi) beside
MPkt. anal.
Lahnda of the Salt Range has 1 sg. ehicd (< *dhiwd), 3 pi. dhin ^ :
^ Bloch, Ulndo-aryeUf p. 243.
“ Hultzsch, CIl. vol. i, Inscr, Aiokay p. Ixxxi.
® lb,, p. xcv.
* See below, p. 810.
^ LSI. viii, i, p. 59.
* LSI. viii, 1, pp. 440, 441.
804
R. L. TURNER —
the same dialect contrasts marin < Pkt. mdrenti with maresan
< Pkt. maressanti.
The existence of the -a- form so early and in so conservative an area
as Shahbazgarhi may be at first sight unexpected. But we have seen
some indication that the plural -enti was replaced by -anti earlier than
the singular -eti by -ati. It is possible, though not demonstrable, that
Asoka Shah, had singular *dks€ti to plural akmriUi.
e cannot demarcate with exactness the areas comprising the two
main developments of Skt. ks.^ In the North-West, from the evidence
of the Shahbazgarhi and other KharosthI inscriptions ® it remained
till a comparatively late period as ks. In the Dardic languages it is
stiU differentiated from Skt. {c)ch : e.g. as cerebral ch opposed to
palatal ch in Pashai, Khowar, Palula, Dameli, Bashkarik, and Shina ;
as ch opposed to dental affricate ch in Gawarbatl ; and as ch opposed
to ch in Tirahl and Kashmiri.
In Maharastrl Prakrit as (c)ch, and in Marathi as s, it has fallen
together with Skt. {c)ch. It may be that this development was proper
to other dialects also : (c)ch forms are more common in Ardhamagadhl
than in SaurasenI,® but these may be due to the greater influence of
Maharastrl on the former. Nevertheless, there is some evidence that
the ancestor of Singhalese, which has some very striking resemblances
with Ardhamagadhl,* was a ch language.^ As for Gujarati, which
I previously « grouped with the kh languages, Bloch points out that
it shares with Marathi a small residuum of words in which ks is repre-
sented by ch (s), and which are not generally found in this form
elsewhere.’ In the time of Asoka the language of Girnar, where
‘ See J. Bloch, Langue maralhe, § 104.
Sten Konow, Cl I. vol. ii, pt. i, p. cx.
’ Pischel, Gr. Pkt. Spr. §§ 317-320.
-tr-, preceded by a long vowel, > -t. which subsequently was lost like original
* t, t- (see . Geiger, A Dictionary of the Singhalese Language, p. xix) : the group
rt(h) > at(h) >al. -a a , e / . o
* W. Geiger, op. cit., p. xxi, and Literalur und Sprache der Singhalesen, p. 42.
« JRAS. 1921, p. 539. 'y ■ c
He instances, op. cit., p. 113. Guj. chudvu [but also khudvu] ‘ to pound ’ = Mar.
sudrie, but Ass. khundiba ‘to pound Hi. khMna ‘ to trample ’ (Skt. ksunatti, Pkt.
chumdm, khumdal) ; Guj. richalvu ‘ to rinse’ = Mar. visalne(ci. Skt. viksalila-). Guj.
- yena — Mar. taras (Skt, taraksa-^ Pk, taraccha-) and us ‘ sugarcane ’ = Mar.
Tri/o' iff** ’ either Mar. or North Guj. where ch > s
li * 7 1 V’ ^ ^ Guj. cAo ‘ plaster, mortar ckoru ‘ to plaster but
eng. kho broken brick Hi. khod ‘ broken brick, mortar ’ {.Skt. ksoda-, ksodati,
iron f-* P’^^'^ier ) ; Guj. Idcho ‘ fomenting or burning feet with a hot
Ncp. Diet cautery ’, Idsg n. spot, discoloration ’ (Skt. laksd-.
SANSKEIT A-KSETI IN MODERN INDO-AEYAN
805
Gujarati is now spoken, certainly shows ks > cch} It has only two
words with {k)kh : ithijhakha (of. Skt. stryadhyaksa-) which with its
assimilation of s and r is certainly an Eastern form of an administrative
term, and samkhitena which as a somewhat technical expression may
also be an Eastern form. That this development was proper to Girnar
is strongly supported by the unique sachaya (Gir., xiv, 5), which
Hultzsch ^ rightly explains as equal to *sathksaya. Elsewhere, even
perhaps in Sanskrit {ksd- = khya-) ks > (k)khy ; and even
Shahbazgarhi, which stiU maintained ks, has kh in sariikhaya. Further,
this implies that ch < ks, since it apparently coincides M-ith ch < ks,
was palatal (a pronunciation borne out by the modern languages,
Marathi, Gujarati and Singhalese in which Skt. {c)ch and ks > {c)ch
fall together) ; and that the South-Western change of ks > {c)ch
was independent of the North-Western ks > cerebral {c)ch, by which
Skt. ks is still distinguished from Skt. (c)ch.
Forms of Indo-Aryan acquired by non-Aryan jungle tribes some-
times retain archaisms which are lost in the neighbouring standard
languages. We have seen the survival of *akhi as dgi in a jungle dialect
on the extreme east of Bengali.^ The survival of a word with ch < ks
in the Bhili dialect of Naika<Ji, which is used in the wildest parts of
the Panch Mahals and Rewakantha immediately west of the Gujarati
area, and which apparently alone among the languages of the West
preserves the archaic dch- * is not without significance as to the
possibility of an earlier and wider extension of ch forms in the neigh-
bouring Indo-Aryan region. In the specimen from Lunawada State,
Rewakantha,® chetar occurs for ‘ field ’. The preservation of r in the
group tr, as in Girnar and still in some dialects of Gujarati, precludes
us from supposing an influence of Maharastri chetla- (> Mar. set)
in which r was assimilated at a much earlier date.
In the East and Centre, but extending into Panjabi, Lahnda,
Sindhi, and some of the Mest Pahari languages, ks > (k)kh. Even
in our earliest MidlA. documents there is much mixture of vocabulary
in this respect ® ; but where ch forms have entered kh dialects as loans
the ch is indistinguishable, as in MarfithI, from original Skt. {c)ch :
‘ Cf. T. Michelson, JAOS. 1910, p. 88, quoted by Bloch, loe. cit.
- ClI. vol. 1 , Inscr. Asoka, p. 26, note 6.
’ See above, p. 799.
* achi ‘ is ’, achati f. ‘ was ’ beside chu. ' am ’, LSI. ix, 3, p. 89.
LSI. ix, 3, p. 89.
* J. Bloch, L’Indo-aryen, p. 81.
806
R. L. TURNER
e.g. Hi. tacJind : 'puchna (Skt. tdksati, prccMti) like Mar. tdstfs : pusne
(contrast Ksh. tachun : prichun).
It appears, then, that Skt. dkseti has its proper phonetic form in
Shah, dksamti, Khar. Doc. (h)acK-, Eish. chuh, MPkt. acchai and Mar.
dsne and perhaps Guj. che, OMaith. akhalu, Bhojp. naikhe, Nagpuria
nakhi, MadhesI naikhi, and EBeng. (Camka) dgi. It is possible that
AMg. acchai also corresponds directly to Skt. akseti. It is remarkable
that, judging from Pischel's collection of forms, ^ acch- did not exist
in SaurasenT (in which ks > kkh) ; it is most common in Maharastrl,
coming second in frequency in Ardhamagadhi.
Among the modern languages the ch forms are to be considered
loanwords in Hindi, Central and Eastern Paharl, Mod. Maithill,
Oriya, Bengali,^ and Assamese ; perhaps in Gujarati and Eastern
Rajasthani. This rather extensive spread of a ch form over kh dialects
is not surprising when we find the cch form already established in tlie
literary languages of Buddhism and Jainism.
As so often with the individual facts of vocabulary in India, we
cannot trace the exact path or paths by which acch- spread from the
M est into this Eastern group. But it occurs twice in the Old Bengali
of the Caryas of Kapha,® as the pres. part, acchante ‘ existing, being .
Since generally the language of the Caryas show double consonants
shortened with compensatory lengthening of the previous vowel.®
the presence of the short vowel with double consonant in this wmrd
(contrast, e.g. ndcaa < Pkt. yacc- and p\ich(tmi < prcchami in No. 6)
suggests that it is a loan-w'ord. And, in fact, it occurs quite frequently
in the ‘ Buddhist Apabhrarhsa ’ of the same school, in the Dohakosa.^
Shahidullah * places their composition at the beginning of the eighth
century a.d., Chatterji ^ at the end of the twelfth. P. C. Bagchi « has
found in Nepal a fragmentary palm-leaf MS. of the Dohakosa of
Saraha dated 220 Nepal samvat = a.d. 1100.
The further extension of ch forms can be traced in more recent times.
1 Gr. Pkt. Spr., § 480.
2 ch forms have not completely driven out the older forms in Bengali as in the
negative substantive verb, see below, p. 810.
® M. Shahidullah, Les Chants mystiques de Kanha et de Saraha, p. 116 (No. ID-
From the other Caryas S. K. Chatterji, Origin and Development of the Bengali
Language, p. 931, quotes 1 sg. acchahu, acchami, 2 sg. acchasi.
* S. K. Chatterji. Bengali Language, p. 1 18.
® Shahidullah, op. cit., pp. 99, 203.
‘ op. cit., p. 28.
' op. cit., p. 119 .
“ Indian Linguistics, v, p. 3.52.
SANSKRIT A-KSETI IN MODERN INDO-ARYAN 807
As we have seen, the Varnaratnakara attests the existence in Maithili
of akhalu till probably the beginning or middle of the fourteenth
century A.Dd By the end of that century or the beginning of the
fifteenth Vidyapati, who was born in the latter half of the fourteenth
century, 2 regularly uses forms of ach-.^ Modern Maithili has ch forms
only.*
In Bhojpuri the substantive verb is bat-, bar- < Skt. vdrtate. But
in North Muzaffapur, on the edge of the Maithili area, forms of ch-
ichi, chd, chath) are also used in the masculine plural.^
Elsewhere the contest between {a)ch- and {a)h- (whether < akh- ®
or < as- < dch- ’’) and the extension of one at the expense of the other
can be observed. From the numerous examples of Bhili dialects
collected by Grierson ® we can estabhsh three areas distinguished
by their forms of the verb ‘ to be ’.
1. The Northern with h-, comprising Magri, Khadak and Kotra
(both in Mewar), Nyar, Vagdi, and Dhar. This forms a continuous
area with the h- dialects of West Kajasthani (Mewari and Marwari).
2. The Central with (d)ch-, comprising Naikdi, Alirajpur, Baria.
Carapi, Ahiri of Kacch, Barel, and Pa\Ti. This forms a bridge between
the ch- areas of East Rajasthani (Jaipurl, Harauti, and part of Malvi)
on the east, and Gujarati on the west.
3. The Southern with (a)h-, comprising Mavci, Nori, Rarii Bhili,
Codhri, Gamti, and Dhodia Konkani. This is contiguous to the (d)h-
area of Khandesi and Marathi.
Situated on the borders of the Northern and Central areas is the
Bhili of Ratlam. Here we find he or chai?
Lying between the Central and Southern areas the dialect of
Rajpipla shows a mixed paradigm : —
Sg. 1 chu or dhe PI. 1 dhe, he
2 che or dhe 2 dhe, he-rd
3 dhe, he 3 dhe, he-rd
* S. K. Chatterji, VarTiaratnakara, p. 1.
^ G. A. Grierson, Introduction to the Maithili Language, pt. ii, p. 34.
* See references in A. F. K. Hoemle and G. A. Grierson, Comp. Diet. Bihdri, s.v.
achh-.
* LSI., V. 2, p. 27.
* G. A. Grierson, Seven Gr. Bihdri, pt. ii, p. 44.
* See below, p. 810. ’’ See below, p. 809.
« LSI. ix, 3, pp. 1-201.
* LSI. ix, 3, p. 35. In the specimen on p. 36 occur 1 sg. maru hd, ni hu, ni u ;
3 sg. tedee hai, karuyo he, karayo che.
LSI. ix, 3, p. 85.
808
R. L. TURNER
The East Rajasthani dialect of NimadI, isolated in the South
between the h- dialects of Blalvi on its north and the (d)A- of Eihandesi
on its south, shows a paradigm in which ah- has invaded the 1 pi. ; —
Sg. 1 clfie PI. 1 dya
2 che 2 cho
3 che 3 che ^
Replacement of a homonym is a circumstance favourable to dialectal
borrowing. Thus SindhI, among other languages, in which rt > t{t),
has hatan^ ‘ to spin ' ( < kart-, cf. Skt. krndtti and Pkt. kattai), but
uses an Eastern form katan'^ ‘ to cut ’ ( < Skt. kdrtati, cf. Pkt. kattai
and kattai). In those languages in which ks > kkh, akseti would become
homonymous with dkhyati ‘ says ’, which does, in fact, survive in
Pa. akkhdti, Pkt. akkh&i, Ksh. (Doda SirajI) dkho ‘ word ’, Pj. dkkhnd
‘to say’, Lah. dkhari. Si. dkhanu, Guj. dkhvu, Bhill and Rajasth.
dkh-, OHi. dkhnd. Conversely, it woidd seem to be not mere chance that
in the East, where on the evidence of dialects in the Biharl and Bengali
areas akkh- < akseti survived, there is now no trace of *akkh- <
dkhyati.
It may now be agreed that a verb, for which in respect both of
meaning and form Skt. akseti provides an acceptable origin, appears
as the substantive and auxiliary verb ‘ to be ’ in its expected dialectal
form in various parts of the Indo-Aryan domain, although one form,
that of the West or South-West, has spread beyond its proper
boundaries. Like the verb ‘ to be ’ in many other languages, it has
been liable to various shortenings which have not affected normal
full words. In OMaith. we have akhalu, not *dkh-, though it is
impossible to say whether this was a shortening of dkh- or an earher
change of *akkh- to akh-. Bhojpuri khe has lost the initial vowel
altogether, like many of the languages with ch forms, such as Ksh.
chuh, Nep. cha, Guj. che. If Bengali still has dche in the present, it has
lost its vowel in the past chila, and even in the present when it is
used as an auxiliary, kari{te) dche > korce.
It has been established that not only vowels, but also consonants,
in inflectional elements, in certain frequently used adjectives and
verbs, in pronouns, and terms of address and postpositions, have
experienced changes not found in normal full words. In inflectional
elements -ss- > -s- and -s- > -A- * ; in the adjectives ‘ big ’, ‘ good ’,
• LSI. ix, 2, pp. 315-16.
2 JRAS. 1927, p. 232 fif.
SANSKRIT i-KSETI IN MODERN INDO-ARYAN
809
‘ all ' double consonants have been shortened, e.g. Panjabi, which
maintains double consonants, has vada, bhald, sab < Pkt. vadda-,
bhalla-, sai'7)a-. In the verb ‘ to say ’ Pj. -kh- > -h-, dhia ^ beside
akhnd ; in the verb ‘to go ' Bhad. -ch- > -h- : gdhnu < gncchati ;
in the verb ‘ to be ’ -t- ( < -tt-) > -r- or disappears altogether : Bhojpurl
bate > bare and bd. In postpositions -jh- and -kh- > -h- in OHi.
mdhi ‘ in ’ < mdjhi (cf. mdjh ‘ middle ) ; OHi. kahn (> Hi. ko)
< *kdkhu. cf. OBeng. kakhii ^ (beside kdkh ' armpit ) ; OGuj. pdhaim
‘ by, near ’ is perhaps < pdkhaim (cf. Guj. pdkh ' side, party ) rather
than OGuj. pdsa7m, which survives as pdse?
Similar developments are certainly to be expected in the verb
‘ to be ’ ; and in Skt. bhdvali we have evidence, not only, according
to J. Wackernagel,^ that ava became o as early as the \ edic period,
but also that in the earliest MidlA. documents the initial consonant
has already lost its occlusion. Pa. hoti, etc.
Bloch ® derives Mar. asne ' to exist, be . from Pkt. acchai ;
but, unlike J. T. Molesworth,® he separates this from dhne ' to be ',
for which together with Si. dh-~' OHi. ah-. Pj. Hi., etc., h-. he
tentatively suggests Skt. dbhavati. But abhavati does not occur in
Pali, and Prakrit according to the Pdiasaddatnaha)f7iava of H. T. Seth
has only the past participle dkua-. Apart from the fact that the
modern languages show no trace of o or m (< ava), it appears unlikely
that a verb so widelv represented in the modern languages should have
left practically no trace in MidlA. Far more likely is it that asne
or *dsne became dhne in the function of the simple substantive
verb. This accords with the difference of meaning as defined bi
J. Stevenson ® : asne ‘ to be usually, continue to be , dh)ie to be
Both forms occur as auxiliaries in the J ndiiesvarl , written in a.d. 1290
but revised later.®
This assumption is rendered still more probable when we consider
^ In view of the phonetic weakness of the verb to say in several languages
this derivation is far more probable than that it is <C Skt. aha.
2 Beanies, Comp. Gr., ii, p. 257; and S. K. Chatterji, Bengali Lar^., p. 761.
^ T. X. Dave, A study of the Gujarati Lavyuage in the Sixteenth Century T.*?.,
p. 161.
* Altindische Grammatik., § 108.
^ Langue marathe., p. 289.
® A Dictionary s Marathi and English, s.v.
^ The root is more properly dh-. The nasalization, when it appears, depends upon
the nasalization of a terminational syllable : thus dhe. : dkid.
* Prin-ciples of Murathee Grammar, pp. 113, 114.
® J. Bloch, op. cit., p. 35.
VOL. vra. PARTS 2 A^'P 3. 52
810
R. L. TURNER —
the kh forms derived from ilcseti. Magahi beside emphatic and archaic
(women’s) 3 pi. halchin ‘ are ’ has unemphatic hahin.^
It may be observed from the English sentences he’s here : he isn't
here ; they’re here : they aren’t here, that in the negative sentences the
verb may carry more stress than in the positive. In Nepali, where an
original monosyllable maintains its length (e.g. so, ko, jo < Skt. so,
yo), *chai (<.acchm) became cha, but remained in the negative
chaina. So in the Bihar! dialect of Kurmall Thar ^ we have nekhe
is not ’ beside dhe ‘ is and Nagpuria ® nakhe beside ahe. There can
be little doubt that in these cases the ah- forms are derived from akh-
forms. It may be further presumed that in languages, which no longer
preserve any trace of kh forms, their ah- is the unemphatic derivative
of earlier &kh-. Traces of this d,h-, not entirely driven out by Western
acch-, are to be seen even in Bengali ; for corresponding to the present
of the positive substantive verb dchi, etc., we have the negative : —
nahi nahi
ndhis naha
nahe nahen *
Lastly, among these languages some, beside dh-, have less emphatic
forms beginning with h-. Thus in East Hindi : Bagheli 3 pi. ahen
beside hal ^ ; West Hindi: Bimdeli 3 sg. dy beside Ae,® Kanauji
1 pi. dhinu beside usual hanu,’’ Banaphari 3 sg. dhai beside ha,i.^ It
is clear that the A-forms are derived from ah-, and that this derivation,
as Bloch suggested, must be extended to those languages, such as
Standard Hindi, which possess only h- forms, e.g. Hi. Ani.®
This {d)h- provides the present, and in some cases the past, of the
substantive and auxiliary verb over the whole of the Central and
North-Western area of India proper, namely Bihari (except Maithili
with ch- and Bhojpuri with hat, bdr-). East and West Hindi, Panjabi,
' G. A. Grierson, Seven Gr. Bihari, pt. iii, p. 32.
^ LSI. V, 2, p. 148.
’ LSI. V, 2, p. 280.
J. U. Anderson, A Hlanual of the Bengali Lang'uane, n 15
“ LSI. vi, p. 22.
' LSI. ix, 1, p. 93.
' LSI. ix, 1, p. 402.
* LSI. ix, 1, p. 483.
» I was clearly wrong in Nep. Diet. a.v. hunu in connecting these forms with Hi.
m I etc., which are < Skt. bhdvati, though the two verbs have exercised
mutual influence on each other. There is much chance of confusion when the two words
differ only m their Towels : Hi. hai ‘ is ’ < dkeeti. hay ‘ may be ’ < bhdvati.
SANSKRIT A-KSETI IN MODERN INDO-ARYAN
811
West Paharl/ Lahnda, Sindhi, and West Eajasthanl. It will be
observed that except for a small enclave in West Pahari (viz.
Bhadrawahi and BhalesI in which Skt. ks > ch) this area coincides
with that to which the change, ks > kkh, belongs.
To sum up, Skt. d-kseti ^ provides the present tense of the sub-
stantive and auxiliary verb over almost the whole domain of Modern
Indo-Aryan. It has in a few regions been supplemented, and here and
there replaced, by mh- ‘ to remain ’ and by descendants of Skt .
vdrtate. Only on the fringes are other verbs employed : in Singhalese
as-, sad-, stJid- ; in Gypsy ® and Dardic * and a little group at the east end
of West Pahari forms of as- or of as-. The isolated Rajasthani group
^ Mandeall, LSI. ix, 4, p. 724 ; Mandeaji Pahap, p. 746 ; Cameall, p. 780 ; Gadi
P- 799 ; Curahl, p. 825 ; Pangriali. p. 851 ; Bhadrawahi and Bhalesi, p. 893 ; Padari,
p. 906.
* Bloch (Ind. Ling., ii, p. 32) show ing that the s of Kashmiri 1 sg. chus is < (a)smi,
supports Grierson’s assumption (Garbe-Festgabe. p. 30) that this tense with its variation
for gender is derived from a past participle, Pkt. acchia- [= Skt. dlsiVa-]. But with
the exception of this and perhaps the Hunza Dom verb referred to below (note 4),
the forms of the present tense everj where seem to be derived from the present tense
of Sanskrit {fiSOS. v, p. 138). Where, as in Maithili, there is differentiation of
gender, it has been introduced secondarily through the influence of the participial
tenses. In the Nepali paradigm : —
Sg. PI.
in.
f.
m.
f.
1
chu
chu
chau
chau
2
chas
ches
ckau
cheu
3
cha
che
chan
chin
the separate feminine forms may be due to a purely phonetic change, whereby cka-.
preceded by the final i of a feminine substantive, adjective, or participle, became che-
(which before a nasal regularly > chi-). The 1 pi. f. cbau (instead of *cheu) is due
to the analogy of the 1 sg. chU, in which there is no a to be influenced by a preceding
-1. The present tense with its special feminine forms,
2 sg. m. carchas < *carad-u (or -o) chas < caramto (or -oo) acchasi
f. carches < *caradi (or -i) (has < caramli (or -ia) acchasi,
provides a marked contrast with the future in -nc. Here in the feminine no -i precedes
the auxiliary ; and consequently, as Rajguru Hemcandra in his Gorkha-bhasd-iydkaran ,
p. 89, specifically points out, there is no change for gender in the auxiliaiy : —
Sg. PI.
m and f. m. and f.
2 garTie chas game chan
3 game cha game chan
“ J. Bloch, Indian Linguistics, ii, pp. 27 If.
* Nevertheless, Lieut. -Col. D. L. R. Lorimer in a letter dated 2nd October, 1935,
informs me that in the language of the Dorns of Hunza, of which the vocabulary ia.
largely Shina or akin to Shina, the present tense of the verb ‘ to be ' is ; —
Sg. 1 cis (or ch-t) PI. 1 co
2 cdi 2 cot
3 m. cd 3 ce
f. a
812 SANSKRIT i-KSETl IN MODERN INDO-ARYAN
with s- in Mewati and Ahirwatl and the Bhfll of Mahikantha, Jhabua.
and the Panchmahals, are perhaps formations from the past tense
which Bloch has shown to be very widely based on the imperfect
of Skt. fl.s-d
The descendants of dkseti appear in their normal phonetic form with
hs in the Shahbazgarhi inscription of Asoka and in the Kharosthi
documents from Niya ; with ch (distinguished from ch < Skt. cch)
in Kashmiri and its closely related dialects of Poguli, Doda Siraji.
and Rambani ; with cch (which falls together with original Skt. cck)
in Prakrit (especially Maharastrl), becoming ch in Gujarati and s in
Marathi ; with kh in isolated parts of Bihar! and the extreme east of
Bengali. If kh forms existed generally in the East, they have been
driven out by ch forms of the West. We were able to trace the process
of replacement in the Maithili area.
Where akh- persisted, we observed the tendency in the unstressed
forms to change -kh- to -h-, which supported the derivation from
ikh- of the {i)h- forms found over the whole of the rest of the ks > kh
area. The Sindhi Shi-, preserving evidence in its paradigm of an original
stem ending in -e-, added cogency to this assumption.
' BSL. x.\xiii, pp. ">5 ff.
Some Dravidian Prefixes ^
By Edwin' H. Tuttle
T)RAHUI regularly has a for ancient short e, and e for ancient weak-
stressed ai. Brahui kane (me) corresponds to a blend of
Tamil dative enalchu and accusative ennai, with the dative-formant
prefixed instead of being suffixed : kane < *kenai. This inversion of
the usual arrangement, parallel with English thereby ~ by that, is one
of the few cases where a prefix is plainly visible in Dravidian. Another
example, explained below, is to be seen in Brahui der (who) for
*er, and in the equivalent Kanara ddrit beside aru and jaru, corre-
sponding to Tamil jar < *€har. With regard to recognized elements of
inflexion and word-formation, Dravidian is nearly always suflixal,
not prefixal : this is why I have ventured to compare Dravidian
with Nubian, which is likewise usually suflixal and only rarely prefixal
{JAOS., 1932, vol. 52, p. 133).
In the American Journal of Philology, 1919. vol. 40, p. 84, I have
shown that initial sn became h in Brahui ; h (from s) in Gondi ; s in
Kui ; t in Malto ; n in Kanara and Tamil ; t in Telugu ; t, s, and h
in the three dialects of Tulu. Afterward, in order to explain a similar
variation of initial sounds, I evolved the theory of an ancient zn,
supposed to make n in the southern tongues ; d in Brahui ; h or zero
in Gondi; s in Kui ; t in Kurukh-Malto {A. J. Ph., 1923, vol. 44,
p. 71). The evidence for the assumed developments was taken mainly
from the following words ; Kanara ndlige, Tamil nd, ndkku, Telugu
ndlike, ndluka, Tulu ndldji, Brahui dui, Kurukh tatid (tongue) ;
Kanara nettar, Telugu netturu, Brahui ditnr (blood) ; southern nir
and nlru, Brahui dir, Gondi ejar. er. jer. jar, Kui siru (water) ; Telugu
noru, Brahui *ddr, Kui sudd, Malto toro (mouth), Gondi mussor (nose).
Brahui *dor is apparentlv represented by Dardic dor (mouth) ; former
contact of Brahui and Dardic is shown by Brahui du — Dardic dui
(hand). Gondi mus- in mussor corresponds to Malto muso (nose), ss
being kept where a simple s should have become h or zero. The ending
of Kui sudd seems to have come from a Kolarian equivalent *moda
or *muda represented in the compound tonwd {JAOS., 1926, vol. 46,
^ In the follovping remarks, as in my other writings on Dravidian, I use j with the
value of Dutch j.
814
E. H. TUTTLE—
p. 310). The ending of Tula ndldji was evidently taken from Tulr
baji (mouth).
Singhalese regularly has d for an ancient initial voiced palata
occlusive, as in dana (knee), diva (tongue) ; the development wa^
probably through dz, dz, 8 . Apparently an older form of Brahui d>o.
such as *d{u or *div, was taken from Singhalese before that tongue
emigrated from Northern India. In constructing the 2 n-theory, I over-
looked the probable source of Brahui dm. The word must be removed
from the list. Contact of Brahui and premigrational Singhalese is
apparently shown by Brahui e (that) and o (that), corresponding to
the equivalent Singhalese e- and 3-.
The 2 W-theory, upset by a belated discovery of Grondi nattur (bloodj.
must be discarded. Such a word as “ hare ” might well travel — with
the thing so named — from Gondi to other Dra vidian tongues ; Kanara
mola (hare), instead of a normal *mosal corresponding to Tamil mujnl
and mu^al, is evidently based on an older form of Gondi nuilol < *niolal.
and Brahui muru likewise looks like a borrowing from the Gondz
word. But it is hard to believe that tribes of hunters could have lost
their word for “ blood ” and then taken it from another Dra vidian
tongue.
In literary Kanara the word adu (that thing) is sometimes prefixed
to an interrogative, as adar (who) for simple dr. This queer construction
is probably derived from the sentence ad en (that [is] what ?), misunder-
stood as a simple word (what), and explains the d that is prefixed to
Brahui der and Kanara ddru (who). It also gives a clue to the
explanation of d in Brahui ditar, dir, *dor.
In most varieties of Dravidian we find at least two simple vowels
used as demonstrative adjectives. Ordinary Gondi has lost the demon-
strative adjectives, and uses pronouns instead, as ad maldl beside
Kanara d mola, Tamil a mnmjal or a mmugal (that hare). We may
assume such a use of pronouns in other varieties of early Dravidian.
Thus it is easy to explain Brahui dir as owing its d to *ad *ir ; compare
English the tolher from that other. I now believe that an initial d in
Brahui nouns, where it corresponds to a southern n, or to southern
zero (if there are any such cases), is a prefix of the kind just stated.
It is also possible that a sentence like *ad *ir (that [is] water) was
mistaken for a simple noun (water) and reduced to dir.
Ancient s has generally disappeared in Southern Dravidian, and
seems to be lost medially in Gondi-Kui and initially in Kurukh-Malto
{A. J. Ph., 1919, vol. 40, p. 83). The lost medial s may be replaced by
SOME DKAVIDIAN PREFIXES
815
hiatus-filling v <w or g <. gw < w. Thus *asan (he) is the basis of
Kurukh as, Gondi on- < *aun < *atean, Kui aan-, avan-, Telugu
van- < *awan, Kanara ava, Tamil avan. Kurukh has added n to
ne (who) by misdivision of *asan *e (he [is] who ?) : when *asan
was reduced to *asa before consonants, the group *asan *e kept n
and was niisdivided as *asa ne. English was formerly encumbered
with sexless genders like those of Latin ; it has lost them, aside from
a feminine applied mainly to artificial objects. We may assume that
early Dravidian possessed sexless genders, and that in nir and perhaps
other such words the n came from the end of a prefixed *asan, the
masculine demonstrative ; compare English a newt from an evete.
Another explanation of prefixed n is possible. Beside the inflected
genitive ena, Tamil has en (my) without any suffix. This form becomes
enn before a vowel. If a similar *enn was formerly used in the other
Dravidian tongues, we might assume that its final n became by
misdivision the initial n of words often combined with “ my”, such
as ndlige, nettar, noru.
Ancient initial sn became s in early Gondi-Kui and t in Kurukh-
Malto. By assuming a reduction of prefixed *asan through *asn
to *sn, or a misdivision of *asti combined with a noun, we have an
explanation for the s of Kui suda, Gondi -sor, and for the t of Kurukh
tatxd, Malto toro. Kui siru seems to represent *asir < Hear, beside
Kanara esaru < Hear (boiled water) ; but *icar is apparently a
compoimd corresponding to Gondi atjdr (boiling water), from attdnd
(boil) and a variant of er not otherwise in use.
In conclusion I would assume that, aside from dul, the words
listed above may have had the basic forms *dlak (tongue), *{dhur
or *udhir (blood), Hhar (water), *or (mouth). The word for tongue
is similar to Kolarian alagg or Idyg. The word for water is similar
to Malay ajer. The word for " blood ” looks like Aryan rudhira.
It is probable that further research would reveal other Dra\’idian
words having consonants prefixed in some way or ways like what I
have suggested. An obvious example is Tamil neruppu (fire) beside
eri (burn).
Joan Josua Ketelaar of Elbing, author of the First
Hindustani Grammar
By J. Ph. Vogel
JN his Linguistic Survey of Indiaf Sir George Grierson has drawn
attention to the first Hindustani grammar, and given some
particulars about its author, Joan Josua Ketelaar, who was a German
in the service of the Dutch East India Compan}’. In the present
paper I wish to supplement the information regarding Ketelaar's
career by means of some biographical data mostly drawn from the
Company’s records preserved in the “ Rijks Archief ” at The Hague.-
Ketelaar’s real family name was Kettler. He was born at Elbing
on the Baltic, 25th December, a.d. 1659, as the eldest son of
the bookbinder Josua Kettler. The future ambassador to the Great
Mogul and to the Shah of Persia started his career in the humble
profession of his father. But while a bookbinder's apprentice, he
grossly misbehaved, robbing his master and even trpng to poison
him. It may be that the master-bookbinder was a disagreeable
person. Anyhow, young Kettler was dismissed, and went off first
to Dantzig, where he committed another theft, and then to Stockholm.
This happened in the year 1680.
Two years later we meet him again at Amsterdam, where he
takes service under the East India Company, which used to draw
a large number of its lower personnel from Germany. Kettler now
becomes “ Ketelaar ”, and it appears that with his name his conduct
also changed. In May, 1682, he sailed to Batavia in the ship 't ^Vapen
van Alhmar, and in 1683 was sent from there to Surat, where he
started as a ‘‘ pennist ” or clerk. Evidently he did well, for he made
quick promotion. In 1687, his chief, Anthony Vogel, who was Senior
Merchant and Deputy Director of the Dutch factory at Surat, made
him “ Assistant ” at fl. 20 p.m.
* L.S.I., vol. ix, Calcutta, 1916, part i, pp. 6-8. Cf. also Pror.. A.H.B., May, 1896.
^ I wish here to give expression to my gratitude for assistance kindly rendered by
Dr. R. Bylsma, Keeper of the State Records, The Hague, Dr, F. \V. Stapel, and
Dr. A. J. Bernet Kempers. To Dr. Stapel I owe most of the information regarding
Ketelaar’s career in India.
818
J. PH. VOGEL —
In 1696 he was promoted to the rank of accountant (“ boek-
houder ”) on a monthly pay of 30 guilders. In this capacity he was
employed first at the Company’s head office at Surat, subsequently
as deputy (“ secunde ”) in the factory at Ahmadabad and, from 1700.
as chief of the factory at Agra. In 1701, on accoimt of his ability,
he was raised to the rank of a Junior Merchant (“ onderkoopman ’ )
at fl. 40 p.m. for a period of five years.
During the years 1705—8 he was twice deputed to Mokka in Arabia,
with the object of purchasing coffee. Notwithstanding great difficulties,
including an encounter with a French pirate, Ketelaar fulfilled this
task to the satisfaction of his superiors. After his return from his
first voyage to Arabia he was promoted to the rank of “ Merchant
on a salary of fl. 65 p.m. This happened by a Kesolution of the
Governor-General in Council, dated 15th December, 1706.
Ketelaar was still away on his second Arabian expedition when
the Central Government at Batavia decided to employ him again
at Surat, “ on account of his experience and capacity in the Moorish
language and customs.” This was on 7th September, 1708. By the
same Resolution he was appointed “ Senior Merchant ” at a monthly
salary of 75 guilders. About this time, the Governor-General and his
Council had projected an Embassy to be sent to the successor of
Aurangzeb, who had died in the preceding year, as soon as the chance
of war would have decided who of his sons was to be that successor.
It was thought that Ketelaar might be usefully employed in this
important mission.
The man first intended to be the Company's ambassador to the
Great Mogul was Cornells Besuyen, the Director of the Surat factory,
and Ketelaar was selected to be his deputy. But when the former
died after a lingering illness, Ketelaar took his place, both as Director
of Surat and as head of the Embassy. The respective Resolution
of the Governor-General in Council is dated 1st August, 1711.
By this time Ketelaar had already started on his expedition,’^
which took him to Lahore, where ^ah ‘Alam Bahadur Shah was
encamped. While the negotiations were still in progress, that emperor
suddenly died on 28th February, 1712 ; the Dutch Ambassador then
became a witness of the series of fights fought under the walls of
'■ An English translation of the journal of Ketelaar’s Embassy was published in
Uie Journal of the Panjah Historical Society, vol. X (1929). An edition of the original
document is in preparation.
JOAN JOSUA KETELAAR OF ELBING
819
Lahore by the four sons of the deceased emperor, and ending with
the victory of the eldest son, Jahandar ^ah. The Dutch Embassy
accompanied the new emperor on his march to Delhi in the hot
season of 1712, and stayed in the capital during the ensuing rains.
All the time the negotiations about the jirtnans, the acquisition of
which was the chief aim of Ketelaar’s mission, were protracted,
and at last, after incredible procrastination, the coveted documents
were obtained.
In October, the Ambassador started on his return journey by way of
Agra, Gwalior, Narwar, Sironj, Sarangpur, Ujjain, Dohad, and Godhra.
On their journey from Surat to Agra the Hollanders had travelled
through Rajputana not by the usual road, but by a shorter route,
which took them through Mewar. The rulers of Rajputana, whose
territories they passed, and especially the Rana of Udaipur facilitated
their progress. It was the high-handedness of an imperial officer
subordinate to the subahdar of Ajmir, which caused serious trouble.
They also had to pay considerable sums on accoimt oi rdhddrl, especially
when crossing the dominions of the Jat chief Churaman.
On their return journey through Central India they met with
much more serious difficulties, and had repeatedly to fight their way
through the peasantry in revolt. They encountered almost every
species of brigands by which the highways of India were then infested — ■
Mewatis, grasias, Bhils, and Kolis. Besides, large sums had to be
spent to secure the necessary escorts from the Rajput chiefs of Malwa,
although on the whole these petty rulers were friendly. The passage
through the mountainous regions of Jhabua and Bariya was attended
with great hardships. When, at last, the Embassy reached Gujarat,
they were greeted with the unpleasant news of Jahandar 's defeat.
The victory of Farrulisiyar meant that all hardships had been sustained
in vain, for it was not to be expected that the new emperor would
acknowledge the privileges granted by his uncle. This was all the
more disappointing, as the expenditure incurred on the Embassy
had been enormous. No wonder that the authorities at Batavia
were little pleased with the negative results of the costly mission.
The blame was laid on the ambassador, although it could not be
denied that Ketelaar had shown throughout a remarkable courage,
tact, and patience.
That this was silently recognized we may infer from the fact that
in 1715 he was entrusted again with an important mission to the
Persian Court. We do not wish here to give an account of Ketelaar’s
820
J. PH. VOGEI, —
Persian embassy.^ It may suffice to state that on this occasion, too,
he exhibited a remarkable ability. But both on the journey from the
coast to Ispahan and during his prolonged stay at the capital he
suffered from ill-health, and after his return he died at Gamroon
(Bandar Abbas) on 12th May, 1718. His remains were buried in
the Dutch cemetery outside the town, and a monument, described
as a “ pyramid, 30 cubits high ”, was erected over his grave. This
monument is no longer extant.^
Ketelaar had bequeathed a large sum of money to the Protestant
churches in his native town, Elbing. One of these churches, named
■■ Zum heiligen Leichnam ”, purchased for the money a new organ,
which was erected against the western wall. The cost was 1,562
florins 29 groschen. An oil-painted portrait of the donor may still
be seen at the side of the organ. It shows a full face, with a straight
nose and resolute chin, covered with a profuse periwig according to
the fashion of the period. In a Latin work, Elbinga Litterata (Elbing.
1742), p. 90, it is recorded that Ketelaar’s nephew, Samuel Griitner,
who had accompanied his uncle on his Persian embassy, and was
the executor of his will, thus became the owner of three large-sized
volumes in which Ketelaar had given his own biography in the Dutch
language. The author calls these volumes “ publica luce, si quid
recte judico, dignissima ” ; but, unfortunately, they are no longer
traceable. It is very much to be hoped that some day they may come
to light again.
A manuscript copy of Ketelaar’s Hindustani grammar is preserved
in the “ Rijks-Archief ” at The Hague. It has the following title ;
“ Instructie off onderwijsinge der Hindoustanse, en Persiaanse Talen,
nevens hare declinatie en conjugatie, als mede vergeleykinge der
hindoustanse med de hollandse maat en gewighten mitsgaders
beduydingh eenieger moorse namen etc. door Joan Josua Ketelaar,
Elbingensem en gecopieert door Isaacq van der Hoeve, van Uytreght.
^ Sir George Grierson has quoted some particulars from the curious account of
a German soldier, named Johann Gottlieb Worms, who belonged to the ambassador’s
bodyguard. It was published with some other writings of the same author at Dresden
in 1737 under the title Ost-lndiaii- uml Per-sianische Reisen by a German pastor,
M. Crispinuh Weisen. A second e<Ution appeared at Leipzig in 1745.
^ The British Consul at Bandar Abbas has informed me that there existed “ a very
old ruin in the shape of a monument situated on the border of the oldest part of the
town {once Gambroon), which was known as ‘ Goor-i-Ferangh ’ {Europeans’ grave),
ut this ruin, and others in close vicinity to it, were demolished about twenty-five
jears ago, when it was decided to build new houses on the site.”
JOAN JOSUA KETELAAR OP ELBING
821
Tot Leckenauw A° 1698.” In English : Instruction or Tuition in
the Hindustani and Persian languages, besides their declension and
conjugation, together with a comparison of the Hindustani with the
Dutch weights and measures, hkewise the significance of sundry
‘ Moorish ’ names, etc., by Joan Josua Ketelaar, Elbingensem, and
copied by Isaac van der Hoeve of Utrecht. At Lucknow, A° 1698.’
We have seen that in 1700 Ketelaar was put in charge of the
Dutch factory at Agra. In a letter dated 14th May of that year the
Director and Council of Surat sent instructions to Ketelaar and his
deputy Isaac van der Hoeve regarding the management of the Agra
factory which was then re-established. From this letter it is evident
that both Ketelaar and his assistant had been employed in those
parts before for a considerable time. This is a point of some importance,
as we may conclude that Ketelaar had acquired his knowledge of
the language not only in Gujarat, but also at Agra and Lucknow,
where Hindustani is spoken in a much purer form. The copy now at
The Hague was written by Ketelaar s assistant at Lucknow in 1698,
and we may perhaps assume that the grammar had been completed
by its author in the same year or shortly before.
The Dutch original was never printed and the manuscript copy
at The Hague is the only one known to exist. A Latin translation
of it was published by David Mill(ius), professor of Oriental languages
in the University of Utrecht, in his Miscellanea Onentalia} It is
through this work that the grammar has become known.
Dr. Suniti Kumar Chatterji ^ has given a detailed account of
Ketelaar s Hindustani grammar in its Latin garb. He arrives at
the conclusion that the Hindustani on which the grammar is based
is not a very pure language, but the patois spoken in the bazars
(bdzdru boll) at Bombay' and Surat. I doubt whether this qualifica-
tion is quite correct. It is perhaps largely' due to the transliteration
of the Hindustani words which is necessarily' unsatisfactory and
clumsy', as it is not based on a .scientific sy'stem. In all probability
Ketelaar could neither read nor write Hindustani ; he had to depend
on hearing and produced the sounds in writing as best as he could,
according to the spelling of the Dutch language. The difficulty
came in when he had to deal with phonetic values which do not exist
in that language, such as the palatal and cerebral consonants.
1 MiU’s Miscellanea Orientnha forms part of his Dissertationes Selectae, pubUshed
at Leyden in 1743.
*Dvivedi Abhinandan Granth, Benares, 1033, pp. 194-203.
822
JOAN JOSUA KETELAAK OF ELBDfG
It does not seem very likely that a man in Ketelaar’s position
derived his knowledge of Hindustani mainly from the bazar. He
had no doubt daily dealings with people of the lower classes, such
as the “ peons ” in the service of the East India Company, but his
work brought him also into contact with respectable Indians belonging
to the commercial caste. These were in the first place the brokers
{daldl) or agents (icakil) who played such an important part in the
trade of the Europeans in India. They were invariably merchants
of good standing, such as Mohan Das whose fame for charity was so
great that his house was spared when Shivaji plundered Surat.
In the course of his mission to the Imperial Court, we see Ketelaar
in touch with the highest dignitaries. In the journal of the Embassy,
it is mentioned that, when paying a visit to the amir-ul umara Zulfiqar
^an, he conversed with the latter in Hindustani, whilst his deputy,
Rogier Beerenaard, made use of Persian. We also find it stated that
the ambassador was on friendly terms with a man like Amanat Khan,
who was the svbahddr first of Gujarat, and later of Malwa.
It occurs to me that the versions of sacred texts reproduced from
Ketelaar’s grammar by Sir George Grierson and Dr. S. K. Chatterji
do not reflect the language of the bazar. How far his Hindustani
has been influenced by Gujarati or by the patois of the Western ports
I am unable to decide. As a first attempt Ketelaar’s “ Instructie ” may
be regarded as a creditable production. Although obviously intended
to serve the practical purposes of trade, it betrays a scholarly curiosity
which is also noticeable in the account of his Indian Embassy.
Altindische und mittelindische Miszellen
Von J. Wackernagel
1. SUBHRTAM BHR-
A LTPERS. ubrtain abaram ist von Tedesco, Zsckr.fiir Indol. 2, 44 ff.,
und von Altheim ebenda 3, 33, zu richtiger Deutung ahnlicher
Awesta-Stellen verwertet worden. (Vgl. auch Lommel, Or. Literatur-
Ztg. 1934, 180, der eine leise Abweicbung zwisclien den beiden Sprachen
in der Verwendung der Phrase feststellt.) Merkwiirdiger Weise hat
keiner der beiden Gelehrten daraiif liingewiesen, dass sich dieselbe
Wendung auch im Indischen findet. Schon der Rigveda bietet 4, 50,
7c, bfhas'pdtim yah subhrtam bibhdrti ,, der den Brhaspati hoch in
Ehren halt “ und 9, 97, 24d rtdm bharat sdbhrtam cdrv induh „ Indu
tragt gut die schone Ordnung Dazu kommt aus der Brahmana-
Prosa PB. 8, 8, 16, ta abruvan „subhrtmn no abhdrfir‘' iti ; tasmdt
saubharam ,, sie sagten : , wohl gepflegt hast du uns ‘ ; daher
der Name saubhara “ ; und die entsprechende Stelle JB. 1, 187,
so ’bravit ,, subhrtam vd imdk prajd abharsam “ iti ; tad eva saubharasya
savbharatvam, subhrtam prajdm bibharti, ya evam veda.
Die tlberlieferung gibt an diesen Stellen das mit su- begmnende
Wort nicht ganz gleichmassig wieder. Im PB. neben dem pluralischen
Akkusativ nah mit der Endung -tarn ; im JB. an der ersten Stelle
neben dem pluralischen irndh prajdh ebenso (allerdings in der
Entstellung zu suvratarn), wahrend an der zweiten Stelle neben dem
Singular prajdm die Handschriften zwischen subhrtam und sabhrtdm
schwanken. Caland schreibt im JB. durchweg subhrtam ; es ware
dann anzunehmen, dass in der Prosa das mit su- beginnende Adjektiv,
das im RV. wie im Iranischen mit dem substantivischen Objekt
kongruierte, erstarrt ware. Man konnte geneigt sein statt dessen
vielme.hr die Kongruenz das Adjektivs auch an den Stellen der
Brahmanas durchzufiihren, also im PB. subhrtd(h), im JB. subhrtd(h)
und subhrtdm zu schreiben. Aber das Schwanken der Endung wieder-
holt sich, wie wir gleich sehen w’crden, anderwarts.
Die Wendung erhalt sich in weiterer vorklassischer Prosa mit
su-sarnbhrta- : Baudh. 2, 6 (p. 43, 8), etdn susarnbhrtdn sarnbhardn
punar eva sarnbharati ; 7, 6 (p. 208, 8), susambhrtam sarnbharanydrn
sambhrtya (Caland, Das rituelk Sutra des Baudhdyana, p. 52).
824
J. WACKERNAGEL
Also mag die altpersisch-awestische Wendung hubrtam bar- zwar
,, une locution traditionelle et religieuse “ sein, wie Benveniste in
Bull. Soc. ling., 31, 63 f., und S. 220 Anm. seiner Neubearbeituni:
vmn Meillets Grammaire du Vieux Perse meint ; aber man hat kauni
einen Grund mit ihm dahinter „ une action mede “ zu v^ermuten.
Vielmehr liegt darin ein \uel alteres den Indern und Iraniern gemein-
sames Erbstiick vor.
Erne Menge solcher den beiden Sprachen gemeinsamen ort-
gruppen und Phrasen ist langst nachgewiesen. Noch neuerdings hat
Lommel (Zschr.fur Indol., 8, 270 ff.) auf den schon von Bartholomae
bemerkten Parallelismus zwischen jAw arsbivd hist»nta „ standen
hilfsbereit “ iind vedischem urdhvo asthat usw. hingewiesen. Ygl. auch
azd{d)-hard, „ Kundmacher “ in Elephantine mp. azdegar (Schaeder,
Iran. Beitr., 1, 66) gegeniiber ai. addhd-hr- das zu der Entsprechung
zwischen gAw. visto . . . azdd : ved. addhd vid- ,, zuverlassig wissen ‘
hinzukommt ; nicht zu gedenken der zahlreichen gemeinsamen
nominalen und verbalen Komposita. (Vgl. upari-syena- unten.)
tJbrigens ist weder die Wendung hubrtam bar- im Iranischen, noch
deren indische Entsprechung im Indischen isoliert. Schon Benveniste
aaO. hat auf das vollig gleichartige altpersische ujrastam (bziv.
ufrastd) pars- hingewiesen, das neben singularischem und pluralischem
Objekdakkusativ ,, griindlich strafen “ bedeutet. Aber auch awestische
ParaUelen sind, allerdings in anderem Sinne, schon beigebracht
worden. So Yt. 10, 68, yal dim hu-irixtam irinaxti ,, wohin er ihn
machtig dahinschiessen lasst “ ; Visp. 14, 1, gd&q, hufranuirdtri
fravmrsninq, hufrdyaMq frdyaezyantqni (ahniich 16, 0 yasnmi . . ■ -am)
,, die Gatha gut hersagend, gut weihend
Aus dem Altindisclien liefert die Brahmana- und die Sutraprosa
weitere Beispiele ; KB. 2, 1 (4, 8 L.) supratyudhdn ahgardn pratyuhet
,, er schiebe die Kohlen gehbrig zmiick “ (vgl. Si^S. 2, 8, 15 [aiigardn]
supratyulhdn pratyuhya) ; Vadhfilas. 59 (ed. Caland, Acta or., 4, 165 f.)
etat supratunriktarn pramicya ,,diesen (Kopf) tuchtig ausgekocht
habend “. Ferner (Caland, Das rituelle Sutra des Baudhdyana 52 u.
Pitrmedh. p. xiii) Baudh. ,Sr. 5, 1, 18 (27, 4) tryangulam . . . susarn-
trptani samtarpya, 6, 25 (185, 16) alhaindm . . . suptarihaddham, upani-
hadhndti ; Baudh. Pitrmedh. 3 (7, 10) cdrusthalirn susambhinndni
bhinatti, 11 (16, 10) etad ddahannm . . . sravohsitam avoksyaund (17, 1)
ai. addha mit kr- i.st bis jetzt anscheinend nur in dem Gana saksdt bezeugt ; nun
dient die iranische Entsprechung zura Beweis dafur, wie echt das Zeugnis des
Ganapatha und wie alt die Verbindung ist.
ALTINDISCHE IIN’D MITTELINDISCHE MISZELLEX
825
athainam susamcitam samcitya ; Hiranyak. Pitrmedh. 3 (36, 4)
kapalani susambhinnani sambhinatti ; 10 (42, 5) saflrdni susampistani
pesayitrd ; vgl. auch 9 (41, 11) susamcitam samcinvanti.
Sogar dem PaH ist diese Ausdrueksform nickt fremd ; z.B. Jat.
1, 222, 26 tarn sugaMtam gahetvd ,, dich fest packend und so melir-
fach in demselben Text.
Yerwandt damit sind die Falle, wo kein substanthdsches Objekt
da stebt, sondern bloss su-tam das Verbum bestimmt : TS. 1, 6, 10, 5
ydtJid vai parjdnyah siiijstam vdrsati ,, wie Parjanya tiicbtig regnet “
(ahnlich TB. 3, 11, 10, 3 ydthd vdi parjdnyah suvrstam vrstvd prajabhyah
sdrvdn Icdmdn sampurdyati). Aus dem Iranischen stebt diesem Beispiel
zur Seite Yast 10, 21 yat cit_ hvast^m anhyeiti „ wenn er gut wirft
Gewiss werden besser belesene Forscber nocb weitere indiscbe,
vielleicbt aucb iraniscbe, Belege aufstobern konnen. Dringender ist
die Frage nach Alter und Herkunft der Wendung. Y as Benveniste
Mem. Soc. ling. 23 (1935), 396 bemerkt : „ Piranien ne possedait pas
d'adverbes tels que bene, male, e5, KanZis, mais seulement des prefixes
hu- dus- . . . Le participe aura done pour seule mission de fournir un
support au prefixe,“ erklart zwar gut, warum man gern nacb der
Wendung griff, lasst aber (abgesehen von der irrtiimlicben
Besebrankung des Ausdrucks auf das Iraniscbe) die Frage offen, wie
man dazu kam, die Wendung so zu formen, und ob ausserbalb des
indoiraniseben Spuren von ibr zu treffen sind. Nocb immer verdient
Beaebtung, was Zubaty IF. 3, 125 ff. (und sebon vorber in einer
1884 in Prag ersebienenen Abbandlung) fiber sie vermutet bat. Er
kniipft sie an eine Form der sogen. ,, figura etymologica “ an, bei der
einem Yerbum oder Nomen verbale der Instrumental eines mit su
zusammengesetzten Yerbalabstraktums aus derselben Wurzel beigeffigt
wird, wie im Rigveda z.B. susamidhd sdm idkire, susamidhd sdnuddhah,
suntdhd nihitah, sunirmdthd nirmathitah, sdnema tdt susanitd sanit-
Tahliih, surued ruedndh. Aucb den Sprficben ist diese Form niebt
fremd. Ygl. etwa TS. 2, 5, 9, 4 suydjdyaja ; K^S. 2, 2, 23 suvimued vi
nnnica ; und besonders, in Rficksiebt auf die Yendung mit subhrtam,
von der wir ausgegangen sind, TS. 1, 1, 2, 2 susambhrtd tvd sdm bhardmi.
Zubaty lasst auf jetzt niebt mebr gangbarem Wege, unter Annabme
von alien Instrumentalen auf -m, unsere Wendung auf solcbe alte
Instrumentale zurfickgeben. Besser wird man sicb damit begnfigen
die Abnlicbkeit und innere Yerwandtschaft dieser Ausdrucksformen
zu betonen, und dabei anzunebmen, dass sie sebliesslieb in grund-
spracblicber Neigung zu derartigem Ausdruck wurzeln. Ygl. fiber
VOL. \UII. PART.S 2 ASD 3 .
826
J. WjICKEEXAGEL
diese ia xdelartigea Wortverbindungen zu Tage tretende Xeigung
(ausser den Ausfiilirungen Zubaty’s) besonders Schulze Quaest. epicae
509 (Nachtrage zu 57 f.) und was Delbriick Vergl. Syntax 1, 256 f.
iiber den ausmalenden Instrumental beim Yerbum vortragt.
(Vergleiche auch Bartholomae Stud. 2, 141 iiber umbrisch subocnn
suboco).^
Daneben sei noch auf RV. 1, 162, 10c hingewiesen ; siihrtd tar
chamitdrah hrnvantu ,, das sollen die Zurichter wohl beschaffen
machen Gemass den oben besprochenen Ausdrucken erwartet
man sulcrtam statt des oxytonierten suljtd ; Grassmann hatte Lust
jenes in den Text zu setzen. Aber Geldner in einer Anmerkung seiner
Ubersetzung meint, dass in dem iiberlieferten suJcrtd vielmehr der
Anfang der merkwiirdigen noch unerklarten Adverbialbildungen aut
-d vorliege, die nach Pan. 5, 4, 58-67 in Ausdrucken mit kr- gebildet
werden.
2. ZuM Jai.miniya Brahmana
t)ber die sprachliche Ausbeute, die dieses altertiimliche Brahmana
liefert, hat teils Caland in den Abhandlungen der Akademie von
Amsterdam, 1915 (,, Yerslagen en Mededeelingen Letterlomde
V, 1, 1915, p. 20 if.) und in den Anmerkungen zu seiner Auswahl
Bericht gegeben, teils Oertel in verschiedenen Abhandlungen, zuletzt
im Journal of Vedic Studies, i (,, Roots and Verb-forms from the
unpublished parts of the Jainiiniya Brahmana “). Es sei mir gestattet.
hier aus demselben Te.xte ein paar einzelne Worter und Formen, die
einer sprachgeschichtlichen Betrachtung rufen, herauszuheben.
2, 37. prdnapdndv utkhvlanli ; sa yah bruydt, prundpanav
udaktrasuh . . . marisyanli . . . sie zerstdren den Aus- und Einhauch.
Wenn einer .sagen wiirde ,, Sie haben den Aus- und Einhauch zerstort,
sie werden sterben . . Caland Auswahl S. 134 hat erkannt, dass in
der unfornilichen 3. pi. udaktvdsuh ein Aorist stecken muss und zwar
ein zu dem unmittelbar vorausgehenden Prasens utkhidanti gehoriger ;
er schreibt udakhdsuh und fiihrt diese Form auf eine Wurzel khd-
zuriick.
Aber wenn es eine solche Wurzel iiberhaupt gegeben hat, kann sie
hier nicht in Betracht kommen. Das im Dhatupatha verzeichnete
‘ Unter den griechischen Belegen der ,, figura etymologica die Lobeck Parcalip»
2, 501 ff. aiiffuhrt, kammt den obigen Fallen am nachsten v^ivovi
bei Euripides (Ion S8'l), unterscheidet sich aber von ihnen dadurch, dass das mil
eu- bcginnende Adjektiv nieht als sogen. pradikatives Attrihut dient, sondern einfach
adjektivische Bestimmung zu vfjLvov^ \st ; ,, wohlklingende Lieder erklingen lassen
ALTIXDISCHE TIXD MITTELINDISCHE MISZELLEX
827
Mai- ist nur in Bhatt. 17, 58, und zwar in der 'Form inodaMidy an, mit
der von Vopadeva gelehrten Bedeutung des Grabens belegt ; .. Sie
gruben aus “ ; also, obwohl aiich bier iid ersclieint, in eineni fiir die
obige JB.-Stelle nicht verwendbaren Sinnc. Ubrigens ist ein solches
prolMinynli ,, grabf aus “ von dem sjTionymen prot-Mmnati nicht
zu trennen. Das junge Priisens Ihayati wud aus dem von der TS.
und MS. an belegten Passiv Midyate von Mum- zuiuclcgeliildot sein ;
vgl. tdy- aus tdyate.
Caland hat das in der U berlieferung gegebene t nicht verwertet,
es gehort vor das s ; offenbar ist der geforderte Aorist zum Prasens
utMiidanti in der Form udaMdtsuh herzustellen. Panini 6. 1. 52 lehrt,
class ,, im Chandas “ bei khid- in Formen mit hochstufiger Wurzel
a statt e und ai eintreten kann. Danach muss man in einem vor-
klassischen Text als s- Aorist von kind- neben ocler statt des von den
indischen Erklarern des Dhatupatha gebildeten ukhaitsam die Form
akhdtsam geradezu erwarten. Solch ein Aorist stimmt zu dem ebenfalls
zu khidati gehbrigen Perfektum cakhdda, das, von der Kasika zu P. 6,
1, 52 gebildet, von Eggeling. Sacred Books 26. 151 Anm. aus SB.
3, 6, 2, 12 (p. 282, 10) a cakhdda . . . dkhidat nachgewiesen und schon
1879 von Saussure Mem. 176 (= Publications scientif. 165) in den
richtigen sprachlichen Zusammenhang gestellt wurde. Aus dem
Kigveda gehort wohl d-cakhdda 6, 61 , Ic auch dahin. Dass der vedische
Xominalstamm -Miada- entsprechend der Lehre Paninis ebenfalls
zu kind- gehort, erkannte vohl zuerst .Aufrecht bei Bohtlingk Wdrkr-
buch Nachtr. 5, 216 s.v. avakhdda-. Auf dieser Erkenntnis haben
claim weitergebaut Fleet JRAS. 1909, 426 f. (im An.schluss an eine
begrifflich zutreffende Feststellung von Keith S. 423) und Charpentier
ebenda 1928, 132.
Erst verhaltnismiissig spat wird khdd- als Hochstufenform von
khid- durch khed- verdrangt. Zwar lehrte einst Roth fiir kkedd im
Rigc'eda, als ob es zu kind- gehore, die Bedeutung ,, Hammer ;
aber diese Deutung des Substantivs ist langst aufgegeben, vgl.
Oldenberg Anm. zu 10, 116, 4 (p. 337) und Geldner. Altester Beleg
von khed- als Hochstufenform von khid- ist anscheinend utkhedana- hei
Baudhay. 21, 5 (3, 78, 1) ; vgl. Caland Baudh. p. 59. (Doch verdient
die Tariante iitkhidana- ernsthafte Beachtung ; Tiefstufe der
M urzelsilbe bei Bildungen auf -ana- neben tiefstufigem Prasens und
unter dessen Einfluss ist von vedischem kirdna- und prkina- neben
kirdti bezw. sprsati an in alien T^’pen des Altindischen zu treffen).
Diese spiiten Bildungen beweisen natiirlich nichts. Es ist unbegreiflich.
828
J. WACKEENAGEL
dass noch in neuesten indogermanischen Etymologika khid- ohne
Beriicksichtignng der Hochstufe khad- mit lat, caedere gleich gesetzt
wird. (Falsch auch Kuiper Acta orient. 12, 200 Anm. 2,)
JB. 2, 238. AbMpratarana ije prajdkdtnas ca khalati ca daivovatt
avacicMtsamdnah ; tato vai so ’va khalati daivovdti (-ti Handschr. !
acchintta ; 239 khalati (drei oder vier Worter) daivovdti avdcchaitsi
[sic r\ ,, Abhipatarana opferte, als er sich NacKkommen wiinschte und
khalati daivovati von sich wegzuschneiden suchte ; darauf schnitt
er Kh.D. von sich weg “ ; 239 ,, ich habe Kh.D. von mir wegge-
schnitten Was ist mit den Akkusativen khalati daivovati gemeint.
die in drei Satzen als Objekte des medialen ava-chid- genannt sind
Caland ubersetzt khalati- mit ,, Kahikopfigkeit obwohl das Wort
nur an zwei spaten von Bohtbngk-Roth v, 1358 aus Sayana
angefiihrten Stellen diesen Sinn hat, sonst immer, von VS. und TS.
an, ,, kahlkopfig “ bedeutet. Unter dem sonst gar nicht belegten
daivovdti- versteht er fragend die Bezeichmmg einer korperlichen
Abnormitat oder Krankheit. Wie man immer von dieser Auffassung
von khalati- urteile, hochst bedenklich ist die diesen Wortdeutungen
K^ylan^^IonTma^eiU
ALTINDISCHE UND MITTELINDISCHE MISZELLEN
829
dh'ovdta- mit i als Endung und Vrddhi der Anfangssilbe, wie z.B.
in KB. daivoddsi-, Patronymikum von vedischem divoddsa-. Ein
Name *divovdta- ist bis jetzt allerdings nicbt belegt, aber durcbaus
denkbar. Er ware Parallelbildung von devdvata-, das im EY. zugleicb
als Mannsname und als Adjektiv in der Bedeutung ,, von den Gottern
begebrt, den Gottern angenehm belegt ist ; als Mannsname liegt
es dem vedischen Adjektiv und Patronymikum daivavdtd- zu Grunde.
Zu diesem devdvdta- verbalt sick der bier angenommene Name
*divovdta- genau wie im Eigveda das Adjektiv dyubJiaJcta- zu dem
Adjektiv devdbhakta-. Mit jenem dyubhalcta- wurden die Ubersetzer
und Lexicographen des Veda nicbt recbt fertig. Die von Rotb ein-
gefiibrte Wiedergabe ,, vom Himmel zugeteilt “ (oder ,, gescbenkt “)
lasst nicbt erkennen, was man dabei unter ,, Himmel “ versteben
soil ; deutlicber kommt dies zum Ausdruck in Geldners Wiedergabe
(zu 1, 73, 6) ,, von den Himmlischen gescbenkt nur dass man nicbt
ersiebt, wie dyu- ,,die Himmliscben “ bedeuten kann ; Ludwigs
,, vom Tage gescbenkt von ratna- 4, 1, 18, fallt wobl ganz ausser
Betracht. Da ,, scbenken “, ,, zuteilen “ einen personlicben Agens
voraussetzen, muss bei dyu- in dyubhakta- an einen wirklicben Gott
gedacht werden, abnlicb wie bei divoddsa- als Mannsnamen und divo
duhitf- (oder duhitf- divdh) als Bezeicbnung der Morgenrote. divah-
als Vorderglied in *divovdta- gegeniiber dyu- in dyubhakta- darf nicbt
befremden. Yon alters ber kann bei Adjektiven auf -ta- der Agens
durch den Genetiv gegeben werden (Delbriick Allind. Syntax 153
u. Vergleich. Syntax 1, 348), und es ist verstandbch, wenn auch dem
Altindischen sonst nicbt gelaufig (Ai. Gramm, ii, 1, 213 [§ 89c Anm.] ),
dass ein solcber Genetiv aucb im Yorderglied eines Kompositums
erscbeint, vgl. griecb. AioaSoTos u. dgl.
Das Adjektiv dyubhakta- und der Name *divovdta- stiitzen sicb
gegenseitig. Sie haben aucb das gemein, dass sie isobert sind, wabrend
die begriffsverwandten Komposita mit deva- als Yorderglied vom
^ eda an sebr beliebt und gebrauchlich sind ; klass. devadatta- ist
sogar typiscb fiir Mannsnamen iiberbaupt. Dies bangt damit
zusammen, dass div- als Gott scbon in der Zeit des Eigveda am
Aussterben ist. Daber ist der in Zee und Ju(piter) fortlebende urindo-
germaniscbe Yokativ den Indern verloren gegangen. Zum alten
Hinunelsgott wurde von den Frommen der vediscben Zeit nicbt mebr
gebetet. Wo die Dicbter eines zu div- geborigen vokathnscben
Ausdrucks bedurften, griffen sie nacb dem ja aucb sonst so ange-
wandten Nominativ, imd gaben diesem vereinzelt den Yokativakzent ;
830
J. WACKERNAGEL—
d(i)yauh neben haufigem dyduh (mit dem alten Beisatz pitar BY. 6, 5] .
5a und AY. 6, 4, 3c ; parallel mit dem Yokativ prthii'i BY. 6, 51.
5a ; 6, 68, 4d ; 10, 59, 8=9=10 ; alleinstebend BY. 8, 100 (89), 12b
(lydur dehi lokdm).
Demgemiiss ist *divordfa- jedenfalls ein sehr altertiimlicher Name.
3 , 66 (bei Caland AusiraJd p. 293 Anm. 15) atharvdna uparisyenatii
svargam lokam dyan und 3, 269 (zweimal) imam ujmrisyetiam svargam
lokam jiglsdma, enthalt auch wieder einen Ausdruck, der zum
gemein.samen indoiranischen Erbgut gebort. Dem altindischen
Ausdruck ,, den oberhalb des Adler(flug)s befindlicben Himmelsraum
entspricht genau jAw. npdirisaena- phlv. apdrsen als Name eine.s
Gebirges. Das Kompositum ist gleicbartig mit uparibudhna-
uparimartya- ,, iiber den Boden “ bzw. „ iiber die Menschen sich
erhebend “ im BY. und mit ttpairizmfna- ,, oberhalb der Erde befind-
lich “ und upairidahyu- ,, der iiber den Landern ist “ im jiingeren
Awesta ; verwandt sind die Avyayibhavas uj/aribhumi und uparimbln
ira Satapathabrahmana.
Das lange d der zweiten Silbe in der iranischen Form kann
natiirlich nicht Yrddhi-Vokal sein, wie Bartholomae Altir. Worterbuch
s.v. meinte ; die ableitende Yrddhi eignet bekanntlich nur der
Anfaugssilbe der Worter. Und richtig hat schon Horn KZ. 38, 292
bemerkt, dass auch die Bedeutimg des Kompositums keine ableitende
Yrddhi erwarten la.sst. Nun wird durch die altindische Entsprechung
das n als iranische Neuerung erwiesen. Langst schon hat Andreas
erkannt, dass in solchen Fallen das a durch e hindurch aus
dem epenthetischen ai entstanden ist. — Gemass der sonstigen
Entsprechimg ahairi : upairi (Berliner Sitzungsber. 1918, 392) bieten
die awestischen Te.xte dieselben Zusammensetzungen mit ahairi-
., unterhalb der Erde, . . . der Lander befindlich “.
3 , 247 . Zweimal hinter einander liest man hier die 1 pi. («« vai
ddre) ’srntna ; eben dasselbe PB. 15, 3, 7. Es ist deutlich eine
Aoristform, da etwas soeben selbst erlebtcs berichtet wird, und zwar
eine solche von sp. Caland Auswahl, p. 287, Anm. 2 fragt, wie die
Form grammatisch zu erklaren sei. Nun klassisch miisste es dafiir
asardma (oder asdrsnui) heissen ; von asardnui (3 sg. asarat) aus ist
aber aspima zum Teil zu verstehen. Die Aoristform asarat folgt in
dor A\ eise der Bildung den 'Wurzeln auf r, mit denen auch drs-
zusammengeht : obwohl der Aorist auf a im allgemeinen Tiefstufe
der Y urzelsilbe fordert, (z.B. alipam, asicatn), ist bei jenen die AYurzel-
silbe guniert und udattiert. Diese Seltsamkeit ist schon langst erklart.
ALTINDISCHE UXD MITTELINDISCHE MISZELLEN
831
(Festschrift Jacobi 16 f., vgl. Whitney § 847 a.) Sie beruht darauf,
dass jene Aoriste mit ar aus alten Wurzel-Aoristen von Wurzeln mit
auslautendem oder inlautendem r herausgewachsen sind, bei denen
ar dem Sincnilar activi, insbes. der 1 sg. eignete z.B. akaram akar.
An diese 1 sg. auf -aram hatte sich mit Ubergang in die sogen.
thematische Flexion eine 2 sg. auf -arah, eine 3 pi. auf -aran usw.
angeschlossen. Ebenso war adarsan nach altem adarsam an Stelle
von vedischem adrsan getreten.
Sonach muss auch als urspriingliche Grrundlage von asarat (das
allerdings schon ini RV. und zwar in augraentlosen Formen mit
Udattierung der Wurzelsilbe belegt ist und vereinzelt als Imperfektum
verstanden worden sein muss) einstigea Dasein eines normal
abstufenden Wurzelaorists mit asarani als 1 sg. und asr- asr- im Dual
und Plural gefordert werden. Demnach batten wir also gegeniiber
dem klassischen asarama eine einstige 1 pi. *asnm zu fordern.
Offenbar liegt erne solche Form unserm aspima zu Grunde. Es ware
bloss noch zu fragen, wieso jene konstruierte Idealform *a$rma um ein
n vor der Personalendung erweitert werden konnte.
Ich glaube, es lasst sich aus dem alten Sprachgebrauch hcraus eine
Erklarung linden. Das konstruierte *aspmt hatte im vorklassischen
^’erbum ganz allein gestanden. MerkwUrdigerweise ist bei keinem
alten Wurzelaorist eines r- Verbums eine 1 plur. auf -rnia belegt,
sondern nur -arma mit hochstufigem ar : vedisch karma, akarma ;
vgl. TS. 3, 2, 5, 4 adarhna gegeniiber JB. adrsma. Hienach ware es
zwar denkbar, dass vorklassisch eine Form *asarma aufgekommen
Ware. Aber der Wunsch nach schwerer Mittelsilbe konnte sich auch
in andrer Weise Luft niachen. Auch das mit sr- sinnverwandte gam-
besass einen alten Wurzelaorist. Und da lautete die 1 plur. vedisch
aganma. An diese Form eines begrifflich verwandten k erbums schloss
man die 1 plur. von sr- an und ersetzte -ma duich -mm.
Auf eine andere Erkliirung ist Caland in der spater als die Auswahl
aus dem Jaiminiyab. veroffentlichten Ubersetzung des Pancavimsa-
brahmana- (p. 394, Anm. 1) verfallen. Ohne auf eine Urform *asr)m
zu rekurrieren. dachte er sich die Form asrmna aus dem Wunsche
entsprungen, den verbalen Ausdruck fur dare sr- ,, in eine Spalte
geraten ‘‘ dem -srt des zugehorigen addrasrl anzupassen, als ob dessen
auslautendes f, u-urzelhaft ware ; man hatte also anzunehmen, dass
-mn- nach der Weise des Satzsandhi fiir -tm- eingetreten ware. Aber
ein zweites Beispiel vmn nni fiir tm im Wortinnern ist aus dem
Altindischen nicht bekannt, und ebensowenig eine solche Einwirkung
832
J. WACKEENAGEL —
einer Nominalform auf eine Verbalform. tJ^brigens ware aach diese
Erkliirung Calands vielleicht verstandlicher, wenn man dem aspimn
ein altes *asrma zu Grunde legen diirfte.
Auffallig ist das sowobl in PB. als in JB, iiberlieferte n statt n .
Man kann es pbonetisch erklaren gemass Ai. Gramm. 1, 187 (§ 167b)
oder aus dem Vorbilde der Musterform aganma hezv,\ des t von
adarasrt.
3. PEAKE. trWUDHA-, UWIpnA-
In einer Maharastri-Strophe der bengalischen Eezension der
Sakuntala best man das AVort uwudha- in der Bedeutung , . aus-
gerissen " (p. 88, 2 ed. Pischel : muha-uwvdha-munalo . . . cakkdo).^
Der indische Erldarer fiihrt das AVort auf Sanskrit udvyudha-
zuriick, leitet es also aus ud-vi-vaJi- ab. Pischel (Granimalik der
Prakritsjyrachen, § 126, p. 100) tritt dieser Deutimg bei ; sie ist aber
unmoglich. Erstens ist die Verbindung ud-vi-vah- dem Indischen
fremd ; sie ist nicht bloss imbelegt, sondern widerstreitet einer der
Regeln, die fiir die Anordnimg der mit einem Verbum oder einem
Nomen verbale verbundenen Praverbien gelten : wenn vi und nd
mit einem Verbum verbxmden werden, geht vi regelmassig voran ;
so schon im RA'’. vyud- mit sthd-, in der TS. mit uh- und hr-. AA eiteres
in der ganzen Literatur.
Dagegen ist die Folge vd-vi nicht bloss in der vorklassischen
Sprache fast unerhort ; aus der epischen und klassischen wiisste ich
nur ud-vjks- nachzuweisen, das sich aus der Engigkeit der Verbindung
erklart, die zwischen vi- und iks- eingetreten ist, sodass vnks- fast als
Simplex empfunden wurde. Dazu kommt aus der jiingeren Sanskrit-
prosa ud-vikdsin- ,, aufbliihend “ in der Kadambari. Nur ein Text
macht eine ausgesprochene Ausnahme, das Bhagavata Purana ;
hier treffen wir Verbindungen wie ud-vi-ghosayati, nd-vi-caks-,
ud-vi-ddrayati, ud-vi-varhana- ; ja sogar 4, 7, 44, doppeltes ud m
ud-vy-ud-asyet. Aber durch seine erkunstelte Sprache scheidet dieses
AA'erk iiberhaupt als Zeuge fiir lebendigen Sprachgebrauch aus.
AAMiterhin ist unersichtlich, wie man von einem ud-vi-vah-, wenn
es ein solches wirkhch gab, die an der Sakuntala -Stelle geforderte
Bedeutung ,, ausgerissen “ herleiten konnte.
Dazu kommt eine dritte Schwierigkeit ; Hemacandra 1, 120 lehrt
^ Pischel (§126, p. 101 oben) weist dieses Wort auch aus der Ardhamagadhl der
Jaina nach. Der Text, den er dafur zitiert, ist mir leider unzuganglich, ebenso wio
dfr eine der Texte, aus denen er § 489 (p. 346) 'UwihaX belegt.
ALTINDISCHE UND MITTELINDISCHE MISZELLEN 833
ir vodmdhe ,, in udvudha kann u za i werden Also gab es eine
Nebenform uvvidha- ; aber diese lasst sich mit keiner Kunst an vah-
anknupfen. Pischel bilft sich hier mit der Annahme, dass un-udha-
und uvvidha- trotz dem Zeugnis Hemacandra’s zwei ganz verschiedene
Worter seien, jedes von anderer Herkunft ; aber wer rvird ihm auf
solch verzweifelten Ausweg folgen ?
Den Weg zum Richtigen weist Pischel selbst durch die Erklarung,
die er fiir uvvidha- gibt. An der von ihm a.a.O. § 126, p. 100 angefiihrten
Stelle Vivahap. 1388 gehort es mit einem Prasens uwihai und einem
Absolutiv uvvihittd zusammen und wird von einem aus dem Kocher
herausgezogenen PfeU (tassa usussa . . . uwidhassa) gebraucht. Pischel
fiihrt dieses uvvidha- uwihai auf ai. vyadh- zuriick (§ 489, p. 346) ;
das ist sichtlich irrig. Das Prasens von vyadh-, das altindisch voni
Rigveda an vidhyati lautet, ist im Prakrit teils durch Pormen mit
vindh- vertreten, die sich aus dem lautverwandten vindati ergaben ;
teils durch solche mit vedh-, zu denen schon das Sanskrit Anhaltspunkte
lieferte. Dagegen fiir Pischels Prasens vihati, das eben nur hinter ud
fingetreten ware, bietet sich als Prasens von vyadh- keine Erklarung.
Insbesondere aber lasst sich die Bedeutung ,, herausziehen “ aus
ud-vyadh- nicht ableiten ; dieses bedeutet im Sanskrit eigentlich
,, in die Hohe schlagen “, das zugehbrige ud-viddha- „ hoch “ ; vgl.
pali ubbedha- ,, Hohe “•
Einfach wird alles, wenn man uwihai uvvidha- auf ai. ud-vrhati
und dessen schon in den Sutras belegtes Adjektiv ud-vrdha- zuriick-
fuhrt. Dieses Verbum bedeutet ,, ausraufen “, wird aber gerade auch
von dem Herausziehen der WafFe aus der Scheide gebraucht, wofiir
ich auf die Worterbiicher verweise. Dass die Yerbindung ud-vrh-
bis ius Mittelindische fortgelebt hat, wird durch pali ubbahati
„ ausreissen “ verbiirgt ; fiir den lautlichen Ubergang von vi aus
vr vergleiche man z.B. A3Ig. vicchuya- aus ai. vrscika-.
Dieses udvrdha- ,, ausgerauft “ passt nun ganz vorziighch auch
als Grundlage von uwudha- „ ausgerissen “, wie es an der Sakuntala-
Stelle gebraucht wird, imd macht weiter das fiir Pischel im Grimde
unerklarliche Schwanken zwischen uwudha- und uvvidha- verstandlich ;
man weiss, wie bimtscheckig die mittelindische Vertretung von ai.
r ist.
Bemerkenswert ist bei beiden Formen die Lange des fiir ai. r
eingetretenen Vokals. Bei uvvidha- erklart sie Pischel aus dem Vorbild
von Adjektiven wie lidha- midha- gegeniiber lih- mih-, eine Erklarung,
die moghch bleibt, auch wenn man uwihai anders als Pischel erklart.
m
ALTIXDISCHE t’XD MITTEIINDISCHE MISZELLEX
Was aber die Form uwudha- betrifft, so hat sie, wenn richtig erklai i
ihre gaaz genauen Entsprechungen in den Paliformen abbUlh'i-
pnribbulha- fur ai. di'rdha- parivrdha-, und diese erklart Bartholoui^
ZDMG. 50, 685, auch in dieser Weise ; nach mulka- rulha- u. (h-
neben miih- riih- sei hulha- zu den ini Pali erhaltenen auf vrlt- zurucl '
gehenden iwA-Formen %vie abbuhya (ai. d-vrkya) und abbuhi hinzugt'
bildet ivorden. Neben diesen beiden unter sich gleichwertigta
Moglichkeiten der Erklarung der Lange darf imnierhin daran erinnei i
werden, dass in der vorklassischen Sprache die erste Silbe von lyrf/e
aus *irMha- gewiss als Lange gerechnet wurde, so gut wie die ioj.
trdha- drdha- usw. {Ai. Gramm. 1, 31 [§ 28], 44 [§ 40], 275 [§ 238a ] ;
Dieses vorklassische fdb- kann sich sehr wohl in mittelindischem -idb-
-udh- fortgesetzt haben.
Wortkundliche Beitrage zur arischen Kulturgeschichte
und Welt-Anschauung. II.
Altindoar. Gobhila-, m.n.pr. ; altpers. \/g{a)uh ,,dicere“.
(Als Zusammenfassung dient der 10. Abschnitt : p. 873.)
Von Walther Wust
T\IE nachfolgende Studie, die ich bereits im Sonderabschnitt zur
,, Vorrede “ meines „ Yergleichenden und etjunologischen
Wbrterbuchs des Alt-Indoarischen “ p. 86 angekundigt babe, verfolgt
einen dreifacben Zweck personlicher und sacblicher Art. Sie will,
zuvdrderst, im Rahmen dieses Festbandes und, wbblverstanden, mit
ibrem ganzen Zubebor ebrerbietige Huldigung sein, dargebracbt
dem grossen, bochverdienten Altmeister indoariscber und indischer
Spracbwissenscbaft, Sir George A. Grierson. Sie will, zweitens, Ergie-
bigkeit und Ricbtigkeit des von mir aufgestellten ,, Anah’tischen Mono-
grapbie-Scbemas “ (des betracbtend-zergUedernden Denkverfahrens fur
wortkundlicbe Einzelveroffentlicbungen) erneut unter Beweis stellen,
eine Absicbt, die icb in Uberlegung und Beispiel scbon an so ziemlicb
alien entscbeidenden Punkten obgenannter Worterbucb- Vorrede
durcbgefiibrt babe. Und diese Studie will scbliesslicb, drittens, das
Gesamtgewebe der so ungewobnlicb wichtigen indoiraniscben Bezie-
bungen um einen weiteren Einscblag vermebren, allerdings nicbt um
einen so bunten, gleissenden, wie sie neuerdings J. Przyluski ^
bereitbalt, sondcrn uin einen von der webecbten Gediegenbeit, die
wir J. IVackernagels vorbildlicber Humanistenband verdanken.
Wackernagels iiberzeugende, das eine Mai mit litauisch-germanischen.
das andere Mai (nach dem Vorgang Bartholomaes) mit rein iranischen
Sprachmitteln arbeitende Deutung der vedtschen Eigennamen Nodhds-
und Pedu-^ bleibt zusammcn mit anderen Aufsatzen desselben
Forschers die beste Erlauterung des Satzes : ,, das Altindiscbe wird
* Ich denke hiebei etwa an folgende Aufsatze des franzosischen Gelchrten : ,, Un
dieu iranien dans I'lnde “ — Rocznik Or. 7, p. 1—9 ; ,, La theorie des Guna — Bull.
School Or. stud. 6, p. 25-3.5 ; „ Les noms de la Grande Deesse ' = Rev. de I’hist.
des rel. 105, p. 185-92 ; ,. La diaspora tyrrhenienne et sa limite orientale “ = Rev.
Hittite et Asianiqiie 3, p. 58-64, laiiter Veroflentliclningen, die ich z.B. zusaramen
mit P. Chantraine, Idg. Jb. 19, p. 168, oder mit Wilh. Printz, Idg. Jb. 16, p. 118, IS,
p. 99., 19, p. 97 fur mindestens sehr gewagt, wenn nicht fur durchaus verfehit ansehe.
* Vgl. ,. Indoiranisches “ p. 405 f.
836
w. wrsT —
wohl noch in manchen Fallen der Erforschung des Awesta Hilfs-
dienste leisten konnen “ ^ (erne Aussage, die man gelegentlich auch
wird umkehren mrissen). Die dabei durchgefuhrte Betracktungsweise
ist nicht nnr kervorragend geeignet, alien geistreicken Attacken zum
Trotz das Feld zu behaupten, sondern sie erweitert auck eine von
Hilka angefiikrte, wissensckaftsgesckicktlick zu enge Ansickt Edward
Schroders, die in der Warnimg gipfeln zu miissen glaubt, ja nicht
,, die Namendeutung als das hochste, oder gar als das nackste ZieJ
der Namenforsckung kinzusteUen. Die grossen Aufgaben auf diesem
Gebiete [seien vielmekr] Gesckicbte der Namenschopfung und der
Namenwakl “
1. Name imd Person des Gobhih, sind innerkalb des altindoariscken
Sckrifttums eine feststebende Grosse. Gobhila ist — und Friedrich
Knauer bat dies in seinen Ausfiikrungen iiber ,, Spracke und Stil "
(2. Heft p. bOimten und f.) unterstricken — der Verfasser des Gobhila-
grhyasutra, das „ als eines der altesten, vollstandigsten und interes-
santesten Grikyawerke bezeichnet werden*"® darf, in verschiedenen
Ausgaben uns heute vorliegt * und, ,, selbstverstandlicb mit Aussckluss
der Spriiche in einem durchwegs klassiscken, die Grammatik nicht
verletzenden Sanskrit (Knauer a.a.O. p. 50) abgefasst ist. Das
Gobkilagrhyasutra gekort zum Verbande des Samaveda, eine Tatsache
die, wie wir weiter unten nock seken werden, fiir seinen Verfasser
selbst nicht ohne Bedeutung ist. Angesichts dieses Sackverkaltes
sollte man meinen, dass der Name des Gobhila auck sprachgesckicht-
kck immer wieder untersucht worden ware, mit anderen Worten, dass
1 J. Wackemagel, Studia Indo-Iranioa. Ehrengabe fur Wilhelm Geiger p. 233 oben.
Alfona Hilka, Beitrage zur Kenntnig der indischen Namengebung. X^ie
altmdischen Personennamen (= Indische Forschungen. 3. Heft;, p. 77, wobei die
Quelle ,, Verhandlungen der 46. Versammlung deutscher Schulmanner und Philologen
in Strassburg 1901 “ (Leipzig 1902), p. 36 ist.
3 Leopold von Sehroeder, Indiens Literatur und Cultur in historischer Entwicklung
(Leipzig 1887), p. 199 A
Ich nenne die Ausgabe „ with commentary by Chandrakanta Tarkalafikara ".
2. Auflage in der ,, Bibliotheca Indica “ 1906-8 erschienen (erste Auflage ebenda
1880), femer die bekanntere Friedrich Knauers, Das Gobhilagrhyasutra heraus-
gegeben und ubersetzt, Leipzig 1885-6, schliesslich die tlbersetzung H. Oldenbergs
in ,, Sacred books of the Ea.st “, Bd. 30. Am iiberaichtlichsten verzeichnet den
Gobhila-Schrifttumskomplex Louis Renou in seiner „ Bibliographie Vedique "
p. <6 f. Fur die Parisigtas, die dem Gobhilagrhyasutra angehangt sind und von denen
eines unter dem Titel Gj-hya.samgrahaparisigta einem Gobhilaputra zugeschrieben
wird, sei auf M. Wintemitz, A histoiy of Indian literature 1, p. 281 verwiesen (vgl.
'Icr indischen Litteratur 1, p. 223, Anm. 2 unten).~Nach Jlonier-
sutra^'p^rf^ Gobhila auch ein Pugpasutra, ein Srautasutra und ein Kaigeya-
AEISCHE KULTURGESCHICHTE
837
aus dem Gesamtverlauf der Indologie und indoarischen Philologie
sich eine gescUossene Erforschungsgescliichte (oder die von mir
so genannte Traditionslinie) liickenlos herausarbeiten liesse. Leider
ist diese Erwartung vollkommen irrig. Ich habe dutzende und aber-
dutzende von Stellen und Werken durchgesehen ohne das geringste
Ergebnis und muss aus Platzmangel darauf verzicbten, die gesamten
Fehlanzeigen bier in extenso vorzufiihren. Eine Stichprobe aus dem
■wichtigsten, insbesondere dem lexikograpbischen, StofFgebiete geniige !
Schon T. H. Colebrooke erwabnt an drei Stellen seiner ,, Jliscellaneous
essays “ (London 1837, und zwar 1, p. 100*, 314 ; 2, p. 8) den Gobhila,
aber die Erwahnung gescbiebt im Rahmen einer blossen Aufzahlung,
ohne den diirftigsten philologischen oder sprachgeschichtlichen Hinweis,
und auch der zuverlassige ,, Index “ fiihrt uns fiir keinen der beiden
Bande weiter. Ebenso verhalt es sich mit dem Bbhtlingk-Roth'schen
PW., welches den Eigennamen ,, eines Verfassers von liturgischen
und grammatischen Sutra “, Gobhila, fiir eine kleine Gruppe uns schon
bekannter Texte belegt (vgl. p. 836*), aber keinerlei Deutung, keinerlei
Nachtrage, Verbesserungen und sonstige Zusatze beisteuert. Das
pw. tritt ihm, gleichfalls schweigend, zurseite, nicht anders Richard
Schmidts ,, Nachtrage “, C. C. Uhlenbecks ,, Kurzgefasstes
etymologisches Worterbuch der Altindischen Sprache “ (gleichfalls ohne
,, Nachtrage und Berichtigimgen “), Ernst und Julius Leumanns
,, Etymologisches Worterbuch der Sanskrit-Sprache “, Lieferung 1.
Die Sonderverbffentlichungen zum Gobhilagrhyasutra verandern
die merkwiirdige Sachlage nicht. Friedrich Knauer hat sich nicht
iiber den Namen Gobhila geaussert, auch Oldenberg nicht in seiner
vorhin genannten Ubersetzung, Bloomfield und v. Bradke nicht in
einschlagigen Aiifsatzen der ZDMG. (35, p. 533 ff. beziehungsweise
36, p. 417 ff.) imd schliesslich nicht die Verfasser der gangbaren
Literaturgeschichten, Bhagavaddatta, H. v. Glasenapp, V. Henry,
A. B. Keith, A. A. Macdonell, L. v. Schroeder, M. Winternitz (deutsche
und englische Bearbeitimg). Hermann Oldenberg in seiner ,, Literatur
des Alten Indien“ erwahnt noch nicht einmal den Namen Gobhila^.
Angesichts eines solchen Tatbestandes wird es voll erklarlich, warum
sogar Louis Renou in seiner reichhaltigen ,, Bibliographie Vedique “
nur eine einzige Deutung verzeichnet (p. 300, Abteilung 194, No. 9),
1 Ebensowenig Richard Pischel in seiner Darstellung ,, Die indische Literatur
( = Die Kultur der Gegenwart. Ihre Entwicklung und ihre Ziele. Hrsgg. von Paul
Hinneberg. Teili. Abteilung vii. Die orientalischen Literaturen. ii. B. 1, p. 160-213),
Berlin und Leipzig 1906.
838 w. wusT —
diejenige M. Bloomfields, auf die ich nachher noch ausfiihrlicli zu
spreclien kommen werde. In AVirkliclikeit hat sich schon mehr als
ein halbes Jahrhundert friiher Lieut. -Col. James Tod, Late Political
Agent to the Western Rajput States, in seinem beriihmten Werke
,, Annals and antiquities of Rajasthan or the Central and Western
Rajput States of India “ mit der Deutung des Wortes GobMla- befasst,
freilich ohne auch nur annahernd vollstandig die Zusammenhange
zu iiberschauen. In dem genannten Werke * namlich, dessen 1. Band
dem englischen Konig Georg IV dediziert worden ist am 20. Juni
1829, schreibt Tod iiber die GoMl’s (1, p. 137 f.), wie folgt ; ,, This
was a distinguished race : it claims to be Suryavansi, and with some
pretension. The first residence of the Gohils was Juna Khergarh,
near the bend of the Luni in Marwar. How long they had been
established here we know not. They took it from one of the aboriginal
Bhil chiefs named Kherwa, and had been in possession of it for twenty
generations when expelled by the Rathors at the end of the twelfth
century “. Die weiteren Entwicklungsstufen, in deren Verlauf
dann noch als geographische Fixpunkte die Namen Saurashtra,
Piramgarh, Bhagwa, Nandod, Si-hor, Bhaunagar, Gogha, Gohilwar
eine Rolle spielen, gehen uns in diesem Zusammenhang hier nichts an.
Fiir uns ist nur von Wert, dass Tod aus vorstehendem Anlass anmer-
kungsweise, 1, p. 137 Anm. 2, den Eigennamen Gohil bezeichnet als ,, a
compound word from goh, ‘ strength ’ ; Ha. ‘ the earth ’ Diese
Deutung, welche nur unter Vorbehalt die ,, zeitlich erste “ genannt
werden darf, well sie ja den fiir uns wesentlichen Zusammenhang mit
dem altindoarischen Personennamen Gobhila- weder erwiihnt noch
erkennt, ist heute nur noch wi.ssenschaftsgeschichtlich lehrreich. In der
Tat gibt ihr Crooke a.a.O. p. 137 Anm. 2 den Zusatz ; ,, [This is out
of the question ; cf. Guhilot.] “, und nicht anders urteilt Irach J. S.
Taraporewala in seinem nachher noch eingehend zu behandelnden
Aufsatz 2 ; ,, The word (ahura) in Sanskrit and the
Gobhilas “, weim er schreibt ; ,, Colonel Tod in his immortal
Rajasthan [Popular edition in two volumes (1914), i, p. 95, ftn. 1.]
derives the word from ‘ goh (strength) and ela (the earth) ’ ; but this
^ Ifh l>enutze die Ausgabe. die ron William Crooke, C.I.E., ,, with an introduction
and notes “ in drei Ilanden (Oxford f-niversity I*res.s 1920) bearbeitet worden ist.
^ In ,, Indo-Iranian studies, being eommemorative papers contributed by
European, American and Indian scholars in honour of Shams-Ul-UUema Dastur
Darab Peshotan Sanjana (London-Leipzig 1925), p. 143-8.— Beames a.a.O. 1,
oV Elliot) sagt kurz und bundig : ,, and his etymologies are not worthy of the
slightest notice “. j o j
ARISCHE KULTUEGESCHICHTE
839
seems rather of the folk-etymology variety “ (a.a.O. p. 147 unten und
f.). Dabei haben weder Crooke noch Taraporewala iiberhaupt auch
nur die Frage gestellt, welche Bedeutung denn dies derart beschaffene
Kompositum besitzen solle — die Antwort wiirde gleichfalls gegen
Tods Analyse ausfallen — und keiner von beiden hat darauf
aufmerksam gemacht, dass Gohil (Gohel) doch in einem Form-Verband
mit ahnlich gebauten Eigennamen zu stehen scheine, wobei etwa die
Rajputana-Personennamen Baghel, Chandel, Jugel, Mohil, Patel,
imd Vadhel hatten genannt werden miissen. Insbesondere die beiden
letzten sind morphologisch eindeutig : Patel gehort zu altindoar.
patta-, m. ,, Tafel, (Urkunden)platte “ (tatsachlich ist auch *pattila-
bezeugt, allerdings nicht als Eigenname), und Vddhd fiihrt auf so
gegenstandliche geschichtliche Vorgange * zuriick, dass die Ableitung
des Namens ,, from badh, radh, ‘ to slay ’ vie schon Tod (2, p. 943)
erkennt, unausweichlich ist. Tods Zerlegung Goh-il besteht also formal
wohl zu Recht, aber -il kann im Wortganzen nicht als selbstiindiges
Nomen, sondern nur als Ableitungssuffix gefa.sst werden, eine Tatsache,
die schon jetzt als sehr wesentlich fiir die folgenden Gedankengange
unter alien Umstanden festgehalten werden muss. Ein bedeutender
Abstand nicht nur zeitlicher, sondern vor allem sachlich-wissenschaft-
licher Art trennt den zweiten Versuch, mit dem Eigennamen-
Komplex Gobhila- fertig zu werden, von dem ersten. Konnte James
Tod nur mit Vorbehalt als Glied der Traditionslinie behandelt werden,
so liegt uns, ganz im Gegensatz dazu, ein vollgiiltiger, sozusagen
dreiteiliger Ansatz vor in dem, was Sir Henry M. Elliot, John Beames
und Fitzedward Hall zur Sache ermittelt haben Da die Einzelangaben
der drei Forscher, soweit unser eigenstes Yerhandluntjsthema.
Gobhila-, nicht beriihrt wird, da und dort im Texte der Untersuchung
selbst untergekommen sind, ist hier unmittelbar nur das zu erortern.
^ ,, Aja [ein Abkoramling der Rathor-Dynastie von Man\ar, Rajputana], another
brother, invaded Okhamandal, in the extreme west of Kathiawar, and established
himself there by murdering the Chavada ruler of the country. His descendants bear
the surname which he assumed, and are still known as Vadhel, ‘the Slayers ’
(Sir Wolseley Haig, The Cambridge Historj* of India 3, p. 521 unten ). — Baghel
und Gohil erscheinen auch in der Gedichtstrophe, unten p. 840b unmittelbar
nebeneinander.
® Es handelt sich dabei um folgende VerotfentiKhung : ,, Memoirs on the history,
folklore, and distribution of the races of the North-Western Provinces of India ;
being an amplified edition of the original Supplemental glossary of Indian terms, by
the late Sir Henry M. Elliot, K.C.B. . . . Edited, revised, and re-arranged by John
Beames, M.R.A.S. In two volumes. Vol. i (London 1869), p. 90-2, drei Seiten, an
deren einer, p. 91 f., „ Dr. Fitzedward Hall, the well-known erudite scholar mass-
geblichen Anted hat.
840 w. wusT —
was Hall den Namen Gobhila- betrelfend denkt. Er steuert folgende
Notiz bei ; ,, It seems probable that the term Gahlot [woriiber unten
p. 849^ f.] is connected with Guhila, the name of the second king of the
Udaypur dynasty. See the ‘ Asiatic Kesearches vol. xvi, pp. 292,
etc. The apparent eponymist of the Gahlots is called Gobhila in an
inscription from Chedi, which I have published and translated in the
‘ Journal of the American Oriental Society vol. vi. Guhila may have
been corrupted from Gobhila, a name occurring in ancient Sanskrit
literature ; or, which is rather more likely, Gobhila may have been
Sanskritized out of Guhila. In popular pronunciation, Guhila must
have become Guhil ; and we know that gotra was broken down into
got . . . .“ Elliot-Beames fahren anschliessend fort : „ In support of
this opinion, I may add, that throughout the Prithvi Haja Rasa,
Chand always uses the form Gohil, So dankens-
wert diese Angaben auch sind, im wesentlichsten Punkte konnen
sie nicht gebilligt werden : die beiden Eigeimamen Guhila- und
Gobhila- haben weder in der einen noch in der anderen Weise etwas
miteinander zu tun. Ganz abgesehen davon, dass dies auch das,
spater noch vorzufiihrende, Urteil Charpentiers ist — vgl. unten
p. 846* — , wird uns ein eigener Untersuchungsabschnitt dariiber
belehren, dass die beiden in Guhila- bezw. Gobhila- zutagetretenden
Wurzeln reinlich geschieden sind Dies ist der einzige, nochmals
aufzugreifende Ansatz, den wir von dem zweiten Deutungsversuche
mitnehmen. Noch weniger fruchtbar kann das ausgestaltet werden,
was Albrecht Weber, Akademische Vorlesungen iiber Indisehe
Literaturgeschichte (zweite, vermehrte Auflage ; Berlin 1876, also
mehrere Jahrzehnte nach Tod), p. 92 unten zum Nonien propriurn
Gobhila- beibringt. Es besteht in der ohne jegliche Nachweise auf-
tretenden Bemerkung ; ,, Sein Namen hat einen sehr unvedischen
Klang, und findet sich durchaus nichts demselben irgendwie
Entsprechendes in der iibrigen vedischen Literatur vor “. Taraporewala
* ,, The following passage from the ‘ Ala udal prastav ’ of that poem is interesting,
as shewing the Gohil in good and valiant company at an early date, in spite of the
later imputation of cowardice. It wdll be seen that they are decidedly on the side of
Parimal, the Chandel king of Mahoba, and opposed to Prithl Raja, though the author
above says their chief, Govind Kao, was an ally of the Chauhan : —
‘ The monarch shouted his commands.
To battle marched the Kawat bands ;
Chandel, Banaphar, heroes leal,
Baghel and Gohil, fierce as steel ’ “.
Man kann sagen : Gnhila- ist von Gobhila- noch mehr versehieden als *gnpila-
von •gopila-. ^
ARISCHE KULTURGESCHICHTE
841
hat, a.a.O. p. 147 u., diese Notiz Webers aufgenommen und glossiert
sie mit den Worten ; ,, The remark of Weber that the name Gobhila
has an ‘ unvedic ring ’ is worth considering Ich kann Taraporewala
keineswegs beipflichten. Bei aller Anerkennung der Weber ’schen
Verdienste muss man doch betonen, dass er im allgemeinen ein
schlechter Wortkundler war und dass seine obige Bemerkung nicht
nur irrig (vgl. p. 861^’ ®), sondern im besonderen unklar ist. Sollte der
Ausdruck ,, unvedischer Klang “ sich auf die Endung beziehen, so
ware auf die mehrfach bereits im Veda belegten Nomina mit dem
Ausgang -ila- hinzuweisen ; sollte er auf ausserindoarische Lehn-
bezugsmbgUchkeiten anspielen, so waren Einzelnachweise zu
fordern ; und die Sehlussworte endlich, es finde ,, sich durchaus nichts
. . . irgendwie Entsprechendes in der iibrigen vedischen Literatur vor
sind so mehrdeutig, dass man sie am zweckmassigsten auf sich beruhen
lasst. Denn wer will entscheiden, ob damit nun Verbalbeziige oder
Nominalbeziige oder beides zugleich oder schliesslich ganz allgemeine
morphologische Beziehungen gemeint sein sollen ! So stosst man.
wenige Jahre nach Weber, auf M. Bloomfield, der bis dahin das
Namensriitsel Gobhila- am entschiedensten anpackt, und zwar in
einer kleinen Mitteilung ,, On the etymology of ^t'Ao? [abstract of
a paper read at a meeting of the University Philological Association,
April 6, 1883] (= The Johns Hopkins University Circulars. Baltimore.
November 1882-October 1883, Baltimore 1883, p. 141 f.) “. Bloomfield
geht, wie ersichtlich, von altgriech. <f)lAos aus und betont : ,, no
less than six distinct etymologies have been proposed for this word,
and none has secured for itself the acceptance of scholars generally.
With the exception of Vanicek, who ventures to accept the
pronominal derivation of the stem (from sva- the reflexive), no larger
work ventures to accept any of the proposed etymologies. Pick's
Comparative Lexicon and the fifth edition of Curtius’s Etymology
are silent on the subject. The latest etymology propounded by Ostholf
in the Morphologische Untersuchungen, Vol. iv, p. 229, according
to which is derived from the preposition (in i^lopKos)
= skr. ahhi, and means ‘ beigehorig, zugethan ’, is based upon wide-
reaching new theories about Indo-European vocalism which are far
from certain. [Absatz.] Thus a suggestion which may secure for the
word a Sanskrit equivalent can be made without attacking any well-
established explanation. There are two proper names in Sanskrit,
Gobhila and Rcbhila, whose first syllable may possibly be the
diphthongal stems go ‘ cow ’ and re ‘ property ’ (Lat. res). In that case
VOL. vrn. PARTS 2 and 3 . 54
842
W. WUST —
the remainder of the words, -hhila, would be sound for sound the
Greek ^lAo- and go-bhih would be ‘ fond of cattle ’ ; re-bhila, ’ fond
of wealth’’. Gobhila is a name which occurs in the Vedic period and n
otherwise totally unexplained “ Bloomfields Deutung ist ein
Musterfall bestcchender, blendender Kombination, und wie alle
derartigen Kombinationen zerrinnt sie einem bei naherer, un
befangener Prufung unter den Handen, Ich will mich nicht dabei
aufhalten, dass der Urheber selbst seine Ankniipfung mit einem
wenig ermutigenden ,, may possibly “ einleitet und dass er uns nur
ein ,, abstract “ hinterlassen hat, bei dem wir das Fehlen genauerer
philologischer Einzeluntersuchungen angelegentlich bedauern. Ent-
scheidend ist Anderes. Entscheidend ist einmal, dass die heutige
wissenschaftsgeschichtliche Beurteilung des altgriechischen nni
keinen Deut von der verschieden ist, die Bloomfield selbst vorfand.
Gewiss gibt es einige Gelehrte, so F. Muller Izn in seinem ,, Grieksch
woordenboek ^ “ s.v. (Groningen — Den Haag 1926), Walde-Pokorny
2, p. 185 und Prellwitz ^ s.v., die noch bis in die neueste Zeit hinein
an der Verbindung des altgriech. <f>iXo? mit irisch bil ,, gut nieder-
land. billijk, mhd. unbilde, n. ,, Unrecht, das Unbegreifliche, Wunder
usw.“ festhalten imd eine idg. Ausgangsform * \/ bhil oder *bhilo-.
mfn. rekonstruieren. Aber schon Boisacq lehnt diesen Zusammenhang
ausdriicklich ab als nicht ,, favorise par la semantique “ und stellt
fest : ,, Et}'m[ologie] obscure “ (auch nichts unter ,, Additions et
corrections “). Und A. 3I[eillet]. dessen Ftihrimg wir uns gerade in
einem Fall wie dem vorliegenden unbedenklich anvertrauen diirfen.
betont in seiner Besprechung des Walde-Pokorny 'schen Wbrterbuches.
BSLP. 28. Comptes rendus, p. 89 : ,, Gr. est sans doute ancien.
Mais il en faudrait determiner la valeur sociale precise, et 1 existence
d'une forme isolee i.-e. *bkilo- est invraisemblable ganz abgesehen
davon dass durch einen Bedeutungsan.satz ,, in gutem Sinne eben-
massig, angemessen, gut. freundlich “ (so Walde-Pokorny a.a.O.)
der Bloomfield'schen Auffassung von altindoar. Gobhila- einiger-
massen der Boden entzogen wird. Tatsache ist und bleibt — - und
deswegen schweigen sich mane he Handbuchverfasser ja fiber (f>iXos
giinzlich aus — , dass das anlautende dank seiner entwicklungs-
geschichtlichen Mannigfaltigkeit zu den zweideutigsten inner-
griechi.schen Lauten gehort (vgl. Boisacq p. xvii unten, Prellwitz -
p. xix unten). eine Tatsache, die ihrerseits wiedcr zur unausbleiblichen
^ Ich habe Bl.'s Meinung ungekurat vorgefuhrt, weil ich glaube, dass nicht jeder-
niann diese eimgermassen entlegene VerofFontlichung zurhand haben wird.
ARISCHE KULTURGESCHICHTE
843
Polge hat, dass jede mit <f>iXos arbeiteude Wortverbindung an dieser
Zweideutigkeit eben teilbat. Es ist desbalb aucb nicbt weiter verwun-
derlicb, dass — ein Knriosum fiir Graezisten ! — E. W. Hopkins,
JAOS. 28, p. 129, altgriechiscb ^iXelv ,, to love “ und (f>lXrma
,, kiss “ eng verbindet mit Eajasthani jndr, Hindi 'piiidr ,, love ”
und ,, kiss “, ,, containing the root of (f>tXrjfia ", womit die Frag-
wiirdigkeit der Bloomfield’schen Gleichung recht kriiftig dargetan
Tvird, und zwar nacb Laut und Bedeutimg, Soviel zum Griechiscben
(vgl. nocb p. 861 f.) Von der altindoariscben Seite her spricht gleich
Mehreres ebenfalls entscheidend gegen Bloomfields Verniutung. Da
es sich hiebei um schlichte Tatsachen handelt, lasse ich sie in der
gebotenen Kiirze sprechen. Erstens bat Bloomfield nicbt bedacht, dass
die relativ spate Bezeugtheit der beiden altindoariscben Eigennamen
Gobhila- und Rebhila- zu dem geforderten erbverwandtschaftlichen
Zusammenhang mit altgriechiscb nicbt sehr glucklich passt.
Enter der Bloomfield'schen Voraussetzung 'ware vielmebi der alteste
Veda, also der Rg\'eda, gerade geeignet genug gewesen, Gobhila- und
Rebhila- zu beherbergen. In Wirklichkeit beberbergt er sie naturlich
nicbt. Zweitens hat Bloomfield mit keinem Worte aucb nur angedeu-
tet, dass das dem altgriechischen entsprechende altindoarische
*bhila- nirgends sonstmehr nachzuweisen ist, ausser eben in den so
zerlegten Eigennamen Go-bhila- und Re-bhila-. Von diesen zwei
einzigen Belegen aber ist — drittens — einer von Bloomfield hand-
greiflich missdeutet worden : Rebhila-, n.pr., das einfach mit dem
funktionell nocb spater genau zu bestimmenden Ableitungsformans
-ila- zu dem bereits iin Rgveda ofter vorkommenden rehhd-, m.
,, Sanger, Sprecher, aucb Eigenname eines Mannes gebbrt. SoUte
Ahnliches etwa aucb fiir den zweiten Beleg Gobhila- zutreffen ? Auf
diese Frage endgiiltig zu antworten, ist bier nocb nicbt der Ort.
Aber wir kbnnen die Antwort wenigstens von der negativen Seite
her schon vorbereiten. indem wir folgende tlberlegung anstellen.
Angenommen die Bloomfield'sche Zerlegung Go-bhila- ware richtig.
Dann wiirde der Eigenname Go-bhila- eo ipso zwei Wortgruppen
angeboren, einer ersten morphologischen, in welcher das Xomen go-
in Verbindung mit Nominibus des Anlauts bh- ° auftrate, einer zweiten,
semasiologiscben, welcbe Wbrter der Bedeutungen Besitzer, Freund,
^ Vgl. noch Franz Dirlmeier, ^cAoy nnd ^lAia im vorhrlleni''tisfhen Grierhcntum,
Munchener Inauguraldissertation 19.^1 ; Paul Kretschmer. IP. 45, p. 267-71 (halt
Boisacqs Zweifcl fur gerechtfertigt und sieht in ^lAo? Lehnwort aus dem klein-
asiatischen Lydischen ; Referat : Glotta 18, p. 238 unten und f.).
844
W. WUST — •
Herr, Meister, Verwandter “ mit dem Yorderglied go- vereinigt zei;
Priifen wir das Material ! Es besteht in der Gruppe 1 aus folgeni; i
Worten : "fgo-bhan^ra-, m. „ ein bestimmter Wasserv'ogel +' -
bhartr-, m. „ Stier “ (Galanos !), go-bhdj-, adj. ,, ein Rind venl -
nend Go-bhdnu-, m. n. pr., go-bhuj-, m. ,, Fiirst, Konig go-bhi
m. ,,fBerg, Eurst, Konig Der formale Befnnd ist eindeutig : 1
jedem der sects Komposita, welche das Nonien go- mit bh- AnschliisM
zeigen, ist das Nachglied unkomponiert entweder Laut fiir Laut odt
in einer durch die Kompositionsregeln geringfiigig veranderte*
Lautgestalt belegbar, und bei jedem der seeks Komposita ist di
vorgenommene Auflosung deshalb aucb zwanglos mogbeh. Ich steli
fest, dass Go-bhila- (und seine Ableitungen) in diesem Form- und Sippee
Verband die einzige Ausnahme ist. In ganz derselben Linie verlaui
die vorhin naher bestimmte semasiologische Untersuebung. D"
Beispiele sind : ^gainm-, m. und ‘\gamhara-, m. „ Besitzer voi
Ktihen gavendra-, m. ,, fBesitzer von Kuhen '\gavesa-, ni
,, Besitzer von Kiihen ^gavesvara-, m. Besitzer von Kiiben
go-pati-, m. ,, Herr der Rinderherden, Anfuhi’er, Herr, f Fiirst, Konig ,
gobandhu-, adj. ,,mit der Kuh verwandt Gobald-, m. n. pr. ernes
Mannes, Gobhatta-, m.n.pr. eines Dichters^, gomant-, adj. ,, Kinder
Kiihe besitzend, reich an Kiihen “, gomanta-, m. ,, fRinderbesitzer
Gomitra-, m.n.pr. eines Mannes, gomin-, m. ,, Besitzer von Bindern
oder Kiihen “, gosakhi-, adj. ,, Kinder besitzend gosvdmin-, m.
,, Besitzer einer Kuh, — von Kiihen “. Selbst wenn man die durch
ihre Bedeutungsentwicklung etwas abriickende Gruppe gavis-, adj.
,, (nach Kiihen) begierig gavisd-, adj. dasselbe, gavisti-, adj.
,, (Kinder) begehrend “ dazunimmt, verandert sich das Gesamtbild
nicht. In jedem der achtzehn Belege ist die in Gestalt eines selb-
standigen Wortes oder unselbstandigen Formans auftretende zweite
Komponente sauber ablosbar und sonst nachweisbar ^ ; nur bei
Go-bhila- ware das nicht moglich. Zusammen mit dem sehon durchge-
sprochenen morphologischen Refund ergibt der gesamte Tatbestand
^ Ich halte die Zerlegung Go-hhatta- fur die natiirliche. An sich liesse sich nn
obigen Gesamtzusammenhang auch die Analyse Gobh-atta- erwagen. Aber gibt es
sichere Beispiele fiir ein derartiges Formans -{a)tta- ? Gekiinstelt schiene mir Annahme
von Haplologie : Gobhatta- < *Gobha-bhatta-.
® Insbesondcre gilt dies fiir goAokhi-, adj. < go-sakhi-y ein Kompositum, das zu
Go-bhila- nur scheinbare und zufallige Parallelitat aufweist. Auch die bei Hilka
a.a.O. p. 120 in der Abteilung „ Pferde- und Rinderbesitz “ aufgezahlten Per-
sonennainen, die mit go-° (und seinen Kasus), "^-gava-y °-gu-y °-gva- zusamtn^n-
gesetzt sind, bieten nichts entfemt zu Gobhila- Passendes. Gobhila- selbst ist gar nicht
aufgefuhrt.
AEISCHE KXJLTURGESCHICHTE
845
wohl eine eindrucksvolle Wakrscheinlichkeit dafiir, dass Gobhila-
eben nicht in die zwei Bestandteile Go + bhila- aufgelost werden
darf. Anhangsweise bemerke ich noch, dass eine genaue Durch-
forschung der re-Komposita zur selben Einsicbt fiihrt. Das
Adjektivnm revdnt- ,, besitzend, reicb, wohlhabend usw.“ ist zu
beurteilen wie obiges gomant-, und Typen wie *re-sakhi- und
Ahnliches feblen iiberhaupt ganzlich. Abschliessend muss somit
festgestellt werden, dass Bloomfields Gobhila-T>eutnug restlos verfehlt
ist, wie namentlich die genaue Priifung des Anklang- und Sippen-
Verbandes der ^o-Komposita gezeigt hat. Bloomfield hat bemerkens-
werterweise den Eigennamen Gobhila- genau so missverstanden.
wie er auch den vedischen Eigennamen Pedu- missverstanden hat
(J. Wackernagel, Indoiranisches p. 406). Brauchbar aus der iibrigen
Darlegung Bloomfields ist nur etwas sozusagen Ausserliches, die durch
ihn zum erstenmal hervorgehobene enge Beziehung zum Eigen-
namen Rebhila-. Hierin liegt ein von mir sogenannter sachlich-
logischer Best, den wir im Laufe der weiteren Untersuchung noch
aufzugreifen haben. Gobhila- aber erweist sich auf Grund der vorher-
gegangenen Erwagimgen als ein Trickbeispiel in der Gruppe der wirk-
lichen p'o-Zusammensetzungen und verbindet sich in dieser Eigenschaft
mit drei anderen Worten, mit : godhiima-. m. ,, Weizen das ja, wie
gerade das Iranische zeigt, gleichfalls ganz andere Vorstufe hat ;
goldttila-, f. ,, ein bestimmtes Tier was vor noch nicht langer Zeit
von J. Charpentier falsch bestimmt worden ist ; *gohira-, n. ,, Fuss-
wurzel das mit unserem Gobhila- wirklich nichts zu tun hat, auch
liber den bekannten -bh-jh- und -r-jl- Austausch nicht, sondern von
Hjalmar Frisk, Zur indoLranischen und griechischen Nominal-
bildung (Goteborg 1934), p. 48 wohl bedingt richtig zu guh
gezogen worden ist. Im Zusammenhang damit gewinnt dann
schliesslich die von mir bis jetzt nicht gebuchte Tatsache an
Bedeutung, dass schon Monier-Williams® Gobhila- nicht unter den
5'o-Kompositis auffiihrt, mit anderen AVorten dass er nicht an eine
Zusammensetzung mit go- geglaubt hat. Alle bisher vorgebrachten
Tatsachen, die der Ermittelung der AA^ahrheit forderlichen wie die
ihr abtraglichen, spielen keine Rolle bei dem nunmehr zu erorternden
fiinften Deutungsversuche, welcher von dem gerade vorhin
genannten schwedischen Indologen Jarl Charpentier herriihrt imd
IF. 29, p. 380 f. zu lesen steht. Charpentier, der — dies sei noch einmal
betont ! — weder die .Ansicht Tods noch die Albrecht AA’^ebers
noch die M. Bloo mfi elds erwahnt, kommt mittelbar auf Gobhila- zu
846
W. WtiST —
sprechen. In der Hauptsache namlich ist es ihm, trie schon die Uber-
schrifb seines kieinen Versuchs zeigt, um „ Pkt. ffo/ia ustt-.“ zu tun.
das er an vier, nock eingehend von mir zu besprechenden, Stellen
des mdoanschen Schrifttums (Desmamanmia ii 89 ; Jacobi, Au.s-
gevrablte ErzaUungen in Maharashtri p. 31, Z. 35 f. ; Mrcchakatikt:
ed. iStenzler p. 31, Zeile 3 von oben ; Devendraganis Vrtti zu
ttaradbyayanasutra iii) aufzeigt und foigendermassen erkiart ;
yMann ist also wahrscheinlich die urspriinglicbe Bedeutung
des, so\nel ich tveiss,nocli nicht gedeuteten AVortes “ (p. 380 unten),
,- Icli erklare ffoM- aus *ffob//a- und fiihre es zu ‘Kub', also
eigenthcb ‘Stier’, dann ‘mannliches Tier ' im allgemeinen, ‘Mann'.
Man vergleiche besonders vrsabhd-, rsabha-, pkt. Usabha-, usaha,
urc tv elches \\ ort goJia~ wohl in seiner Bildung beeinflusst tvorden
stelle ich weiter ein Paar ai. Eigennamen, namlich
(jobMa-, N. des bekannten Sutraverfassers (Gobhila- : goha
- Paksila- :paksa-, usw.) . . .“ i (p. 380 unten und L). Es ist festzu-
stellen, dass diese Darlegungen einiges Brauclibare und Wichtige
enthalten, was sich fiir die Gesamtheit des Untersuchungsganges
\ertverten lasst. leh rechne hieher, dass Charpentier als erster
inner a b der ganzen Traditionslinie Prakrit goha- nebst den dazu-
gehorigen Literaturbelegen herangezogen imd altindoar. Gobhila-
zu goha- m ein bestimmtes Form-Verhaltnis gebracht hat, wobei
go a augenfallig in seiner Rolle als Ausgangs- und zugleich
Mittelstuek hervortritt. Ich rechne weiter hieher, dass durch die
Proportion Gobhila- : goha- = paksa- : Paksila- gleichfalls zum
erstennial das Bildungsprinzip des Eigennamens Gobhila- angegeben,
Venn auch durchaus nicht schiirfer bestiinmt tvorden ist. Aber diesen
uauchbaren Euizeltatsachen stehen grobe Mangel gegcniiber, die viel
nachdruckhcher ins Gewicht fallen und unbedingt hatten vermieden
r en iniissen und kbnnen. Merkwiirdigertveise hat Charpentier
selber teilweise auf diese schwachen Punkte hingewiesen, indem er
a.a^ Einwande gegen seine eigene Deutung erhebt
und unbereinigt stehen la.sst ; „ Es scheint offenbar, dass *go-bha-
- P- Fursten,
sein sollte “ ' Tt* K • " eigentlich mit Gohhila- identisch
Wal l t"": f'f ^as lex. Wort guhila- n.
schlr4 re^^ ‘verborgen- gehort. E.s ist namlieh
(a.a.O p 38in ( h T " gehoren soli -
Kurzfor^'u XaiJ:Te T. tat-ehlioh" vird GuA,7a-
■‘UBgesprochen haT. - siehe uoten ! -
ARISCHE KULTURGESCHICHTE
847
unter Anlelmung an vrsa-bha-, rsa-bJia- zustandegekomnien ist, da es
sonst eine Bildung voraussetzte, fiir die kaum ein zweites Beispiel
hervorgezogen werden kdnnte [sic !]. Denn wir kaben ja das Suffix
-bho- in Tiernamen binter -n- Stammen, z.B. vrsa-hlia-, eXa(j>os usw.,
hinter -i- Stammen, z.B. epi^os , und -o- Stiinimen, z.B. gardabJia-,
daneben finden si<jb Worte, wo die Vorsilbe ^\■u^zelllaft aussielit, s.
Brugmann Grundr.^ ii, 1, 388 f. Fiir eine Bildung *g"riii-bko- fehlt
aber in anderen Spracben jede Stiitze [sic !], sie muss also auf speziell
indiscbem Boden entstanden sein. Es ware interessant zu wissen,
ob die Prakrt-Spracben nocb andere derartige Bildungen ins Leben
gerufen baben ; augenblicklicb sind mir keine bekannt [sic !] “. Eines
Kommentars bediiifen die vorstebenden Satze, welcbe von einer
geradezu unbegreiflicben Sorglosigkeit zeugen, wobi kaum.
Charpentier batte wabrbaftig besser daran getan, auf die
Etjunologisierung des mittelindoariscben goha- ganz zu verzicbten
und sicb damit zu bcgniigen, statt dessen scblicbt auf die Moglicb-
keit eines Zusammenhangs mit altindoarisch Gobhiln- aufmerksam
zu macben. Denn bewiesen bat Cbarpentier diesen Zusammenbang
ebensowenig. wie er die Ausgangsform *g'‘'~>ubho- bewiesen bat, ein
ricljjiiges Wortungetiim vergewaltigender. konstruktiv-tbeoretiscber
Spracbbetrachtung, das icb biemit ein fiir allemal auf sicb beruben
lasse 1. Aber damit nicbt genug, bat sicb Cbarpentier aucb nocb andere
^ Charpentier hat sich noch nicht einmal die doch wirklioh naheliegende Frage
vorgclegt : falls -t'Ao- in deni postulierten *g^'0HbhO' nuht das Tiernamenformans
ist, kann es dann nicht die in der XorainalkomjKisition gern venvendete Ablautstufe
~hha- der altindoarischen -^^bhd ,, glanzen, scheinen. strahlen ■' sein, ivie z.B. in
altindoarisch °-dbha-, „ ahnlieh ”, oder °-iiibha-, ,. ahniich, gleich “ ? Fiir *gohha-
---- inittelindoarisch gaha- crgabe sich unter dieser \'orau.ssetzung die Bedeutung
., wie das Rind (die Kuh, der Sticr) aussehend ”. Ich habe an Charpentiers Stelle
die Priifung dieser Frage mir vorgenommcii und vollige Ergebnislosigkeit ermittelt.
Frstens iveist der ganze Xominal-Verband der \/ bhd samt ihren zahlreichen
Fraepositionalbildungen gar niciits Entsprechendes auf, insbesondere niehts
mit dem zu fordernden, sekundaren Ausgaiig -i/u- (wegen Gobkila-), und zweitens
Hndet sich auch in der doch viel umfangreicheren allgcuieinen Gruppe der -ila- Nomina,
wie wir noch sehen werden, ebcnfalls kein cinziger Parallelfall. Da also ortformen
wie *°-dbhila-, *°-nibhila- nicht vorkomnien, ware es unsinnig, Gobhila- < *gobha-
< *go — bha (< .y/ bha) erklaren zu wollen. Nimmt man diesen Refund mit dem
vome gegen Tod ermittelten (p. 838 f.) zusanimen, so ergibt sich emeut die Notwendig-
heit, -bh- in Gobkila- zur Lautung Go- und nicht zum Ausgang -ila- zu ziehen. Durch
die writer unten vorzunehracnde Prufung des Silbengrenze-t erbandes wird uns
diese Notwendigkeit erhartet werden. Ich mache noch darauf aufmerksam, dass
die rgvedischen -bha- Formans-Bcispielc (vgl. Grassmann. tl'orterbuch Sp. 1709.
Kolonne 4 von links) saratlich dreisilbig sind und dass die sonstigen -bha- Belege
(Grassmann a.a.O. Sp. 1699, Kolonne 2 von links) lauter verbale -bh- Ausgange
zeigen. Auch dieser Tatbestand spricht entschieden gegen *guoubho-.
848
vr. wusT —
Xachlassigkeiten zuschiilden kommen lassen, die ihn ihier^n
gehindert haben, der WahTheit entscheidead nahe zu kommen. L
formuliere die wesentlichen Verstosse kurz und bezeichne damit el>
soviele wesentliche Untersuchungsaufgaben, denen wir zum
nachher noch zu geniigen haben : erstens hat Charpentier k' ii
kritische Behandlung der bis zu seiner Zeit vorliegenden Deutuns'
versuche zum altindoarischen Eigennamen gegeben, ein Versaumni-
das im vorstehenden 1 . Abschnitt meiner eigenen Studie ausgegliche
wird ; zweitens hat Charpentier die Ausgangsform zu altindoarisc;
Gobhila- wie auch mittelindoarisch goha-, *gobJia-, falsch und ohn'
geniigende Unterlagen als go-bha- < *g'^oubho- analysiert, wie ic*.
bereits dargetan habe ; insbesondere fehlt — drittens — jeglichei
Beweis dafiir, dass dieses Nomen *go-bha- niorphologisch durc!'
(v)rsabha- usw. massgebend beeinflusst worden sei ; die Ursache dieses
Versagens ist — ■ viertens — unschwer darin zu erkennen, das'
Charpentier uberhaupt keine scharfe, saubere Analyse der Grund-
bedeutung und der Bedeutungen des Nomens goha- versucht hat, nicht
hat versuchen konnen, -weil er den Stellen-Verband keiner vertiefteii
Betrachtung wtirdigte ; und all dies war schliesslich — fiinftens
bedauerlicherweise unausbleiblich, weil Charpentier, in unbekiimifter-
tem Drauflosetjunologisieren der junggrammatischen Schule, erne
rein individualistisch-subjektive, atoraisierende Betrachtungsweise
pflegte, statt sich in die geheimen Lebensregehr der Einzelsprache
und ihrer assoziativen Gemeinschaftsverbande sorgsam einzufiihlen.
Gegeniiber dem so zu kennzeichnenden fiinften Deutungsversuch
tritt der sechste und letzte mit weit weniger Anspriichen aut.
stammt von Taraporewala, ist oben bei Besprechung Tods und
Webers bereits teilweise herangczogen worden und besagt (a.a.O.
p. 147 unten und f.) : ,, We begin at once to think of the famous
Rajput clan — the Gohils — in connection with this name [Gobhila-]-
There seems hardly any doubt that the two names are identical. . . ■
The meaning of the word gobhila itself is doubtful, and no guess could
be ventured as to its etjunology. The legend of their founder Goha,
the ancestor of Bappa Rawal, is evidently another attempt at popular
etymology to connect the name Goha with the sanskritic word guha
(cave) [Op. cit., pp. 180 ff.J Zu dieser Darlegung ist zu sagen, dass
sie wissenschaftsgeschichtlich genau so von dem durch Taraporewala
anscheinend unabhangig gemachten, gliicklichen Fund des Zusammen-
hangs zwischen dem Rajputen-Namen Gohil und dem altindoarischen
Ramen Gobhila- beherrscht wird, wie in Charpentiers Versuch der
AEISCHE KULTCHGESCHICHTE
849
Hinweis auf mittelindoarisch goha- als brauelibarste Feststellung
hervortritt. Im einzelnen wird zwar dieser Fund Gohil ~ Gobhila-
nocb einiger Bestatigung bediirfen, aber die Gleichung ist, auch auf-
grund des p. 839 f. Vorgetragenen, so augenfallig, dass ich fiir die
weitere Untersucbung damit fest rechnen werde. Dass Taraporewala
die Bedeutung des Wortes Gobkila selbst dann fiir zweifelhaft eracbtet
und binsicbtlich der Deutung keine Vermutung vorschnell wagt, ist im
Rabmen des von ihm Gewollten als weiterer Gewinn zu buchen wie
nicht anders auch seine kritische Haltung gegeniiber der Tod’schen
Fjrklarung des Eigennamens Goha. Hierin ist ihm, wie im Fall Gohil
selbst, William Crooke mit gesundem Urteil vorausgegangen, indem
er a.a.O. p. 259^' Tods Auffassung, Goha bedeute soviel wie ,, cave-
bom “ (a.a.O. 1, p. 259), unnachsichtig als ,, folk-etymology “
bezeichnet
2. Der erste Teil des diachronistischen Untersuchungsweges ist
damit zu Ende gegangen. Wir haben den altindoarischen Eigen-
namen Gobhila- durch das abendlandisch-nordamerikanisch-
kolonialenglische Fachschrifttum hindurch verfolgt, die Traditionslinie
durch Ausscheiden alles irgendwie Unrichtigen und Wertlosen sauber
herausprapariert und daneben ein paar richtig-wertvolle Tatsachen
gefunden. Diese Tatsachen aber waren bedauerlicherweise von so
vielen unscharfen Gedankenelementen durchsetzt, dass noch eine
Reihe besonderer Untersuchungen sich als notwendig erwies, solche
des Sach-Verbandes (zum Rajputen-Namen Gohil), des Silben-
grenze-Verbandes, des Bedeutimgs- und Stellen-Verbandes (zu
den Wortformen mittelindoarisch goha-, altindoarisch Gobhila-),
schhesslich solche des Klang- und Form-Verbandes (-ila- Bildungen.
insbesondere Rebhila-, -y/ gah). Erst wenn diese Sonderpriifungen
* In das GuhilotlGrahilot-Froblem lasse ich mich hier nicht ein, da mieh diese
Erorterung zu weit fuhren wurde. Ich verweise statt dessen auf Crooke bei Tod
1, p. 259^, wo an Ableitung beider JJamen vom Herrschemamen Guha oder Guhasena
(559-567 a. D.) gedacht ist. Ausserdem belehrt uns Tod 1. p. 260' selber, dass die
alte Bezeichnung Gohil, nicht Gehlot (durch Umschrift bedingte Seitenform zu den
obigen GuhilotjGrahiloi) sei. Auch aus diesem Grunde erweist sich eine Erorterung
des Problems als nicht vordringlich. Sehliesslich sei noch auf EUiot-Beames
a.a.O. 1, p. 90-2 aufmerksam gemacht, die von einer Form = Gahlot
ausgehen und niitzliche geschichtliche Einzelheiten bringen, aber in der sprach-
wissensehaftlichen Erlauterung des Namens sicher in die Irre gehen. Oenn weder
die Ableitung von dem Namen eines Sklavenmadchens, Gahla, noch die von Beames
befiirwortete Herkunft ,, from a form Guhilavant, or ‘ Guhila's people ’ “ wird
heute mehr anerkannt werden, ganz zu schweigen von dem volksetymologischen
Versuch, den Hohleimamen zu Ehren zu bringen. Ich selber denke an Guhilaputra-,
n. pr., das ja tatsachlich belegt ist (vgl. pw. s.v. Guhila-).
850
W. WUST —
befriedigead geklart sind, brauchen wir nieht aaebr von den von mir
so genannten sachlich-Iogischen Resten zu sprechen, die sich aus der
Erforschungsgescbicbte des Eigennamens Gobhila- so mannigfaltig
entwickeln liessen.
3 . Stellen wir zu diesem Zwecke die Vorfrage, ob uns die Fort-
setzerin der diachronistischen Gesamtlinie, die einbeimisch-indiscbe
Spracbwissenschaft, nicht dock bei der Bereinigung dieser sacMicb-
logischen Reste behilflich sein konne, so ist mit einem runden Nein
zu antworten. Soviel ich sehe, spielt die Frage des Namens Gobhila-
bei Nairuktas und Vaiyakaranas gar keine RoUe, und dies aus einer
ganz verstandlichen Sachlage beraus ; man fand fiir Gobhila- nicht
den traditionell notwendigen Anhalt an einer Verbal vrarzel Es ist
desv'egen auch erklarlich, warum das Mittel-Indoarische nicht weiter
Xotiz von dem Namen nimmt, eine Tatsache, die etwa durch das
Verhalten des Pali beispielhaft belegt wird Ungefordert, aber auch
ungehindert stehen wir damit vor den nun folgenden synchronistiscben
Betrachtungsweisen, die uns dem endgultigen Ziele zufiihren soUen.
4 . AIs erster Aufgabenkreis ist sinngemass in Angriff zu nehmen
der Sach-Yerband, der den Gohils angeschlossen ist und seinerseits
wiederum in mehrere Teilaufgaben zerfiillt. Leider fliesst das Material
sehr sparlich. Eine Reihe sonst ausgezeichneter Einzel- und Sammel-
werke versagen vbllig ^ und Sri Muni Jinvijayas Studie ,, Kathiyawar
adi ke Gohil “ (erschienen in der Zeitschrift ,, Nagari “ 13, No. 4,
ilagh, Sam. 1989 und angezeigt in ,, Annual bibliography of Indian
archaeology for the year 1933", p. 84, No. 358), die vermutlich tiefer
^ Ich brauohe lu fliesem Zusaminenhang nieht eigens zu betonen, class mit der
(iranimatikerw urzei *hhil ,, bhedane " (dazu die Formen bhilati, bkelayati) umao
^^eI)iger auzufangen ist, als y'' bhil mit go- so gut wie nichts zu schaffen hat.
^ Man Tergleiche die einschlagigen \\'orterl>ucher .sowie Edward Muller, Pali proper
names (- Journal of the Pali Text .Soc. 188H, p. 1-107, besonders p. 29 f.).
,, The Imperial Gazetteer of Iiulia, \ ol. xxi. Pushkar to Salween Xew edition
(Oxford 1908) enthalt nichts^ Meder etwas untcr Goha, noch etwas zu dem Stich-
wort noch auch etwas in dem grossen Artikel Rajputana. Ebenso
\ersagt Sir Athelstane Baines. Ethnography (castes and tribes) (~ Grundriss der
Indo-Arischen Plnlologie und Altertuniskunde li. Ikb, .a. Heft); der Name Gohil
Mird auch nicht im ,, Caste index" (p, iJH-IOo) erwahnt. Vergeblich sucht man
sohliesslifh Belchrung bei Sir Herbert Kisley, The people of India (Calcutta-London
1908), bei H. H. \\il&on. Glossary of judicial and revenue terms and of useful words
occurring in official documents . . . (London lSo4), bei Hermann Goetz, Epochen
der indischen Kultur (Leipzig 1929), bei K. P. Jayaswal, Histoiy of India 150 a.d.
to 3o0 A.o. (Lahore 19.33), m der sonst vorzuglichen „ Enciclopedia Italiana in
rebgion and ethics “ (Edinburgh 1908-1926). Wie
em Geiehrter, der rait wichtigen oder mindestens nicht unwichtigen
en Schaftigt ist, bekennen dasa er nahezu allein auf sich gestellt ist .’
ARISCHE KXJLTUBGESCHICHTE
851
in die ganze Sache eingefiihrt hatte, ist mir leider trotz alien
Bemiiliungen nicht zuganglich gewesen. Nichtsdestoweniger ist,
namentlich aucli wenn wir das vorne p. 838 f. aus Tod 1, p. 137 f. und
p. 839 f. aus Elliot-Beames(-Hall) Angefiilirte hier einbauen, eine
einigermassen geschlossene Tatsachenlinie skizzierbar. Die Gohils
sind ein hervorragendes Geschlecht, das seinen Ursprung vom
Sonnengott Surya selbst ableitet und das in verschiedeneu, von Tod
a.a.O. naher bezeichneten Landschaften Rajputanas und iiberhaupt
Vorderindiens eine wicbtige Rolle gespielt bat . . . und noch beutigen-
tages spielt. Denn ,, tbe cbief . . . [z.B. des Native state
Bhavnager oder Bbaunagar in the Kathiawar agency, Bombay ‘'J
is head of the famous clan of the Gohel Rajputs of Kathiawar. The
Gohel Rajputs are said to have settled in the district about 1260 ‘‘ ■■ a.d.
ilit diesem Doppel-Datum kommen wir zu dem zweiten. wichtigen
zeitlichen Stutzpunkt, zu dem kurzen, \del zu kurzen Satze
Sir Wolseley Haigs in „ The Cambridge History of India " 3, p. 521
oben, wonach ,, in Marwar communities of Gohels, Chauhans, and
Pawars disputed the authority of the Pratiharas or Parihars Leider
gibt Sir Wolseley Haig keine genaue Jahreszahl, aber unmittelbar
vorher nennt er in anderem Zusamnienhang .. 1156 n. Chr. ; das
ganze 20. Kapitel zudem, dem die Stelle entnommen ist, behandelt
,. The native states of Northern India from a.d. 1000 to 1526 ”,
sodass wir mithilfe schliesslich der Gesamttafel ,, Chronology (z.B.
p. 666 0 ., 667 f.) in eine geschichtliche Teillage xms hineinversetzt
sehen, welche durch die Jahreszahlen 1128, 1210, 1212 A.D. ungefahr
abgegrenzt wird. Zwischen das Rajatum der Gegenwart und das
Clantum des 12. und 13. Jahrhunderts n. Chr. aber schiebt sich eine
bisher vbllig iibersehene Tatsache ein, welche nicht nur literargeschicht-
lich Sanskrit und Prakrit rniteinander verbindet, sondern auch zwischen
altindoarisch Gobhila-. mittelindoarisch goha- und neuindoarisch Gohil
die bisher fehlende Briicke schlagt. Diese Tatsache erscheint in eineni
Variantenvers des fiir Edelleute bestimmten epischen Gedichtes
Vlracaritra des Ananta (A'or 1400 n. Chr.), ist uns durch H. Jacobis
verdienstliche Ausgabe in den ,, Indischen Studien “ 14 (pp. 97-160,
genau a.a.O. p. 99 zuganglich gemacht worden und stammt letzten
Endes von dem Jaina Bhavaratna. dem Verfasser des Konimentars
' Einzige Angabe aus ,, The Encyclopaedia Britannica. Fourteenth edition. 1929 “
3, p. 494, Sp. 2, die unter dem .Stichwort Gohel selbst gar nichts bringt. Zeitlich
etwas spater ist der Inhalt der Gedichtstrophe anzusetzen, die p. 840' in Cbersetzung
wicdergegeben ist. Konig Prthviraja i.st 1193 a. D. gefallen.
852
W. WUST —
SukhabodMka (1712 n. Chr.). Der Variantenvers, der, wie gesagt,
eine Zufiigung Bhavaratnas ist, lautet :
„ Yudhisthiro ^hhut kila rajavamsajah,
sa rdjaputrah Paratndravarhsahhuk |
sri. Vikramdrko, nanu Sdlivahano
Gohillabhur vai, Yijaydbhinandandh |
Sisodardnvayabkavo bhavisyati,
tato Nagdrjmia samjnako nrpah
Es ist klar, dass dieser zu einem mittelalterlichen Texte spat
bezeugte Vers mit genealogisch-kosmogoniscben, ja geradezu rein
mythengescbichtlichen Vorstellungen spielt aber es ist sprachge-
scbichtbcb ebenso klar, dass im sonst nirgends mehr belegten
Eigennamen Gohilla- inmitten altindoarischer Wortstaffage em
Prakritismus auftritt, dass nach guten Regeln der Wortbildungslehre
Gohilla- = Gohhila- ist, dass zu beiden mittelindoarisch gdha-, m.
die Ausgangsform bildet 2 , und dass durch obigen Vers aucb die sacbliche
Bindung mit dem ebenfalls Laut fur Laut gleichen Rajputennamen
Gohil gewonnen ist. Mit diesem Ergebnis ist ein grosser Schritt vorwarts
getan. Sofort stellt sich uns aber folgende Erwagung in den Weg :
das im Sanskrit auftretende Prakrit-Lehnwort Gohilla- stebt wobl
zweifellos mit einem fiirstlichen Stammbaum in engem Zusammenhang
' So im wesentlichen nach A. Weber, ZDilG. 24, p. 398 f., besonders p. 399',
wo kurz auf Elliot, Memoirs on the Xorth-Western provinces of India 1, p. 91. 92
(ed. Beames) aufmerksam gemacht ist ; p. 839 und p. 839* f. ; Webers Lesarten und
lesartenvermutungen sind ubrigens in Jacobis Abdruck grossenteils beriicksichtigt.
Eine Ubersetzungswiedergabe des sprachlich ziemlich einfachen Verses erubrigt
sich wohl. Pber den Eigennamen Gohilla- nur das Notigste im pw. s.v. (nichts
im PW. und in Kichard Schmidts Xachtragen). — Zum VIracaritra bringen lediglich
das Landlaufige ; Victor Henry, Les littcratures de I’lnde (Paris 1904), p. 250 f. ;
A. Bemcdale Keith, A history of Sanskrit literature (Oxford 1928), p. 292 unten ;
M. Wintcmitz, Geschichte der indischen Litteratur 3, p. 342 f., 342* (ohne ,, Nachtrkge
und Verbesserungen “),
* Pischel, Grammatik der Prakrit -Sprachen § 595 (p. 402-4) : ,, Dialektisch
uberaus hautig ist -ilia, das fiir -ild steht (§ 194) “. AIs Beispiel nenne ich aus vielen ;
ganthilla- = granthila-. Uber das Suffix -ilia- handelt auch Alfred C. Woolner, Intro-
duction to Prakrit* (in den ,, Panjab University Oriental Publications “), Calcutta
1928, p. 77 unten und f. Ihm zufolgc ist -ilia- ublich in Mahara^tri, Jaina-Maharastri
und Ardha-Magadhi. An Beispielen bringt er kesarilla- (zu kesara-), bdhirilla-
,, external gdmilla- ,, peasant puvrilla- ,, previous Wiederum ist an der
Gleichung Gohil Gohilla- = Gobkila- ebensowenig ZU zweifeln wie an der
eituug dieser Xamendreiheit aus einem vorauszusetzenden *gobha-.
ARISCHE KULTTJRGESCHICHTE
853
und das mit ihm identische Eajputenwort GoJiil bezeichnet ebenso
zweifellos ein Adelsgeschlecbt hohen Eangs, aber ist es angesichts
eines so klaren Sachverbalts denn nicht kulturgeschichtbch haltlos,
beide Wortformen mit Gobhila- zu verbinden, dem altindoariscben
Eigennamen, der zwar spracblich semen Fortsetzern genauestens
gleicht, aber docb eben einen Brahmanen, einen Priester, einen schrift-
stellernden Tbeologen meint ? So merkwiirdig es klingt ; der Wider-
spruch, der in dieser Frage liegt, ist nur scheinbar, er ist auflosbar im
Lichte klarerhellter kulturgeschichtlicber Vorgange. Die Gohils waren,
wde uns Tod erzablt, ,, a distinguished race ; it claims to be
Suryavansi, and with some pretension “ und ,, the ancient Gohils
‘ of the land of KJier expelled and driven to Gohilvral, have lost sight
of their ancestry “ (Tod 1, p. 266 unten und f.) Diese drei Umstande,
aus denen ich den mit dem Stichwort ,, Suryavansi “ besonders
hervorhebe, riicken die Moglichkeit sehr nahe, dass die Gohils
(Gohillas) samt ihrem Stammvater Goka urspriinglich iiberhaupt keine
Ksatriyas und Eajanyas waren, sondern Angehorige des Brahmanen-
standes, ein Ubergang, der uns in der Tat fiir mehrere Eajputen-
Clans einwandfrei bezeugt ist. Crooke erzahlt namlich in der
,, Introduction “ des 1. Tod’schen Bandes p. xxxiii f. folgende lehrreichen
Hergange : ,, Hence arose the legend . . which describes how, by
a solemn act of purification or initiation, under the superintendence of
one of the ancient Vedic Eishis [sic !] or inspired saints, the ‘ fire-bom ’
septs were created to help the Brahmans in repressing Buddhism,
' Aus diesem Grunde ware es auch ein vergebliches Bemiihen, die Lehrerliste des
zum Sama-Veda gehorigen Vamsa-Brahmana heranzuzieben, auf die nach dem
Vorgang Albrecht Webers Taraporewala a.a.O. p. 147 f. aufmerksam macht. Diese
Liste zahlt insgesamt 59 Lehrer auf, die von Brahman Svayambhu selhst angefuhrt
werden und als ersten menschlichen Vertreter Kasyapa haben. Der 26. Lehrer in
Deszendenz von diesem ist Radha Gautama, der seinerseits zwei Schuler hat ; einer
davon ist Gobhila. Vgl. noch Max Muller, A history of ancient Sanskrit literature
(London-Edinburgh 1859), p. 436 f. und p. 442 unten. So wenig wie diese Lehrerliste
fuhrt auch die von Crooke bei Tod 1, p. xxxiv f. und von Tod 1, p. 259 f. selbst erzahlte
Goha-Sage welter, welche die tjbertragung der Konigswiirde von den Bhils an den
Stammvater Goha berichtet . Wer annehmen wollte, dies legendare Ereignis habe eine
Spur in dem Namen Gobhila- hinterlassen, der wurde schnell durch die Uberlie-
ferungsgeschicht e des Bhil-Namens widerlegt. Denn alt und richtig bezeugt ist
eben nur die Form Bhilla- (vgl. PW., pw., Schmidts Nachtrage, Hobson-Jobson
8.V.). Damit scheidet auch die schon in anderem Zusammenhang beruhrte Moglichkeit
einer Haplologie. Annahme einer Form *Gobha-bhila-, aus, ganz abgesehen
von all den gewichtigen Griinden, die bisher schon fiir eine Auflosung Gobh-ila-
gesprochen haben. Auch der zeitliche Abstand spricht gegen die Annahme einer
dcrartigen Beziehung zwischen Gobhila-, Onha und dem Namen der HhTl.‘<. da die
Herrschaftsubemahrae nach allem. was wir wissen, wcsentlich spater als der vedische
Name Gobhila- anzusetzen ist.
854
W. B UST
Jainism, or other heresies, and in establishing the ancient traditional
Hindu social policy . . . This privilege was, we are told, confined to four
septs, known as Agnikula, or ‘ fire-born ’, the Pramar, Parihar,
Chalukya or Solanki, and the Chauhan Ahnlich berichtet Crooke
,, Introduction “ 1, p. xxxiv f. : ,, The actual conqueror of Chitor,
Bapa or Bappa, is said in inscriptions to have belonged to the
branch known as Nagar, or ‘ City ’ Brahmans which has its present
headquarters at the town of Vadnagar in the Baroda state. This
conversion of a Brahman into a Rajput is at first sight startling, but
the fact implies that the institution of caste, as we observe it, was then
only imperfectly established, and there was no difficulty in believing
that a Brahman could be ancestor of a princely house which now
claims descent from the sun Man erinnere sich, dass die Gohils,
nach der Feststellung Sir M olseley Haigs, zusammen mit den Chauhan
im Kampf gegen die Parihars standen und dass der Heros eponymos
der Gahlots nach inschriftlichem Zeugnis Gobhila- war. Dann wird
der Schluss unvermeidlich, dass der Ahnherr der Gohils und Gohillas
ein Samaveda-Theologe Gobhila ( < *Gobha- > Goha-) dnrchaus gewesen
sein kann.
5 . Die Mbglichkeit eines solch lehrreichen kulturgeschichtlichen
Zusammenhangs lasst sich aber auch auf sprachgeschichtlich bedeu-
tungsmassige M eise dartun durch eine Untersuchung des Stellen-
\ erbandes, den Charpentier nur oberflachlich gepriift hat. Und
diese Untersuchung wird, bildlich gesprochen, zum Briickenbogen
werden, der vom sicheren zeitlichen Standort der Gohil-Ans'atze
im 12. und 13. Jahrhundert n. Chr. (p. 851 unten) liber die einzehien
Datierungspunkte hinweg vorstossen wird in den ungestiitzten
Raum spatvedischen Altertunis, dem wir andererseits den Eigen-
namen Gobhila- zuweisen mussen. Damit zur Sache ! Das mittel-
indoarische Xomen goha-, ni. komrnt an 4 Stellen des indoarischen
Schrifttums vor, wie hier nochmals wiederholt sei ; in der
Desinamamala des Hemacandra, in der Kommentarerzahlung
,, Udayana “ zum jinistischen Uttarajjhayana, in der Vrtti des
Devendragani sowie im II. Akte der Mrcchakatika. Eine unmittelbare
Yerbindung mit der geschichtlichen Atmosphare der Gohils wird
dadurch hergestellt, dass drei dieser Zeugnisse dem gleichen Zeitraum
angehoren: Hemacandra hat von 1088-1172 n. Chr. gelebt und
Devendragani ausgangs des 11. und anfangs des 12. Jahrhunderts
gewirkt. Zu den 4 Stellenbelegen insgesamt ist Folgendes zu
Iittelmdoarisch goha-, m. erscheint zunachst Desinamamala
AEISCHE KULTURGESCHICHTE
855
ii 89^, in der Pischel’schen Ausgabe p. 102 oben (ohne ,, Corrections
and Additions “) :
,, gdhammi gdhutl gdyarigod gaggarle a
gdmapahdne gdmanigdmaudaggdtnagohagohd ya“, ein Vers, der in
dem uns angebenden Stiick erlautert wird : „ gdmanl gdmaiido gdtnagoho
goho etc cdtvaro ^pi grdmapradhdndrthdk j goho bhata ityanye | puru^a
ityelce | Aus dieser Stelle und ihrer Glossierung lernen wir, dass
Charpentier irrefuhrend als einzige Bedeutung nur die Kommentierung
durch bhata-, m. und als Ubersetzung damit nur etwa ,, Soldat,
Soldner “ angegeben bat, wabrend die so belangreicbe dritte
Bedeutung — goha- mit drei SjmonjTnen zusammen = ,, Haupt-
person des Dorfes, (Dorf)vorsteber “ — einfacb unterdruckt worden
ist. Die Bedeutungsfixierung „puru§a ityehe“, also etwa ..Mann,
Menscb “ wirkt, wie scbon jetzt festgestellt werden kann, reichlich
allgemein. Das zweite Vorkommen des mittelindoariscben Wortes
goha-, m. ist in Hermann Jacobis ,, Ausgewablten Erzablungen in
Mabarashtri “ p. 31, Zeile 35 f. aus der Kommentargescbicbte
,, Udajana “ zum Uttarajjbavana gebucbt. Das Satzstuck lautet :
„esa rdyd tdva manui jriyd, anne ya gohd"^, was Charpentier unter
Zuhilfenabme von Laksmivallabhas Kommentierung ,.apare mattidydh
ke ^pi rdjdno na santi “ (a.a.O. p. 380^) ganz zutreffend ubersetzt :
,, Dieser Konig bier ist ja mein Vater, die anderen (Konige) sind lauter
Soldaten “ Ohne dass ich mich weiter in die Einzelbeiten der reichlich
verschachtelten Geschicbte einzulassen brauche. kann einfacb festgestellt
werden, dass die Bedeutung ,, Soldner, Soldat “ bier der bereits
behandelten zweiten Angabe der Desinamamala angemessen ist.
An der dritten Stelle des goha- Vorkommens ist das aber bereits
nicbt mehr der Fall. Es handelt sich hiebei. wie scbon angegeben,
um die von Devendragani verfasste Vrtti zu dem Uttarajjhayana,
III. Kapitel, die von Charpentier a,a,0. p. 380- ganz kurz ohne
^ Ich benutze ,, The Desinamamala of Hcraachandra Edited with critical
notes, a glossary and a historical introduction by R. Pischel and G. Buhler. Part I.
Text and critical notes, by Pischel (= Bombay Sanskrit Series. Xo. xvii.). Bombay
1880. Der Variant enapparat enthalt nichts Bemerkenswertes.
- Jacobi a.a.O. ,, Worterbuch p. 107 , Sp. 1 unten gibt das Landlaufige :
,, goha (DK. 2, 89 goho bhatah) Soldat 31, 36
® Ahnlich John Jacob Meyer, Hindu tales. An English translation of Jacobi's
Ausgewahlte Erzahlungen in Mahara.shtrl (London 1909), p. 108 : Again she
reflected : ‘ I’ll enjoy the objects of enjoyment. As regards this king now, he is mv
father ; and the others are simply soldiers ' Oder, wie Meyer m einer Anmerkung
noch beifiigt, ,, his soldiers (sonst keine Bemerkuiigen und auch keinc ,, Corrections
and additions "). Tatsachiich heisst es spater (p. 109 unten) : Then in the month
of Jettha, Udayana hurriedly took the field together with the ten kings [his vassals]
856
w. m^sT —
Weiterungen erwahnt tmd von Jacobi, ZDMG. 38, p. 2 unten und ff.
abgedruckt worden ist. Der Text best ; ,,goha sivardvehim bheravan/
karenti : Jacobi iibersetzt richtig „dieDiener machen einen schiecken-
erregenden (Larm) durch (Nachahmung von) Scbackalgeheul “ (a.a.O.
p. 4 unten), und einc Randglosse in der Handschrift B. erlautert den
Ausdruck gohd (nom. plur. masc.) durch ,, subhatdh was ebenfalls
soviel wie ,, Soldling, Soldat “ besagt. Letzten Endes kann diese
Angabe aber nicht stimmen, denn die vom Fiirsten ausgesandten
gohd werden im Zusammenhang der Erzahlung ganze 3 Zeilen vorher
schlecht und recht purisd genarmt und man sieht schwer ein, waruni
das gleiche Wort nicht auch sofort nachher hatte verwendet werden
konnen. Da dies tatsachlich eben nicht geschehen ist, konnen goha- und
purisa- nicht vbllig synonym sein. Und nun gewinnt aufeinmal die
Satzangabe ein verandertes Gesicht : die gohd erscheinen da ja als
menschliche Wesen, welche — -vielleicht durch Tauschung oder Ver-
kleidung — Larm, Radau hervorbringen. SoUte hierin etwa die purisa-
von goha- abriickende Bedeutungsschattierung stecken ? Diese
entscheidende Frage ist, soviel ich sehe, mit Ja zu beantworten,
sobald wir uns der vierten und altesten goha- Stelle zuwenden. Es ist
der 4. Auftritt des II. Aktes in der Mrcchakatika. Die Szene ist wohl
so bekannt, dass ich bloss mit ganz knappen Stricken zu skizzieren
brauche. Der Bader, „ der zehn Goldstucke schuldig ist hat
Reissaus genommen, verfolgt von dem Bankhalter und einem Spieler,
und ist in einen Tempel gefliichtet. Hier postiert er sich als Statue
hin und wird im schweigsamen Halbdunkel des leeren Gdttertempels
nicht erkannt, obwohl Bankhalter, und Spieler an dem vermeintlichen
hblzernen Bilde riitteln. Der Verfolgung miide, setzen sich Bankhalter
und Spieler, wie wir annehmen diirfen, und spielen verschiedene
Spiele. Der arme Bader, welcher natiirlich in seiner selbstange-
nommenen Maske alles mitbeobachtet, wird erneut von der Spiellei-
denschaft gepackt, und, wahrend die beiden anderen wegen der
Reihenfolge in einen kurzen Wortwechsel geraten, lasst er die Maske
fallen, bricht das muhsame Schweigen, tritt von der anderen Seite
rasch heran und spricht ; ,, Sollte ich nicht dran sein ? “ Darauf
der Spieler ; ,, laidhe gohe ! Charpentier iibersetzt den Ausrufesatz :
„ der Kerlist gefunden“. In der Tat ist dies der ungefahre Sinn der
Situation, weswegen auch die einheimischen Kommentare sowohl
wie die modernen Ubersetzer nicht anders formulierenh aber es ist
11 eine Auswahl des
‘ deatet gohe mit purnmh ;
Wissenswertesten. Das Suvarnilamkarana zur
ebenso die Mrcehakatika-Ausgabe* von Kaknath
AEISCHE KULTTJRGESCHICHTE
857
doch eben nur der ungefabre Sinn. In W'irklicbkeit aber wird der
goJia- aucb in dieser kostlichen Szene als ein mannlicbes, menschlicbes
Wesen gefasst, das sprechend Verkleidung und Schweigen bricht.
,, Bursche, Kerl, Mann “ als Wiedergabe fiir ,,go}ia“ ist folglich zn
blass, zu aUgemein, zu schwach. Ich schlage vor, diesem Erfordernis
durch eine Ubersetzung ,, da haben wir den Sprecker (Larmmacber.
Radaumaclier) ! “ zu geniigen, ohne dass freilich diese deutsche
Ubertragung alle Anspriiche befriedigte. Mittelindoarisch goha-.
m. bedeutet also „ (in Aufputz, Verkleidung) auftretender Sprecher
eine Feststellung, die umso nachdriickliclier wirkt, als sie durck eine
klar analysierbare Textaussage des 4.-5. nackckristlicken Jakrkun-
derts gedeckt wird und bisker von niemandem erkannt worden ist.
Dies Ergebnis wird bestatigt und fortgefuhrt durck den Stickwort-
artikel goha- in Pandit Hargovind Das Sheths Paia-Sadda-Mahanpavo.
A comprehensive Prakrit-Hindi dictionary with Sanskrit equivalents,
quotations and complete references (Calcutta 1928) 2, p. 381, Sp. 1,
der folgendermassen lautet : ,, goha pum [de] 1 gariiva ka mukkiya ;
(de 2, 86) I 2 bhata, subhata, yoddha ; (de 2, 86 ; maka) | 3 jara,
upapati; (upapr 215) | 4 sipahl, pulisa ; (upapr335) | 5 purusa, adami,
manusya; (mrccha 57) | ‘k Dazu treten die Zusatze in Band 4 (Pari-
&ta), p. 1240, Sp. 1 unten : ,, goha pum [de] 1 kotavala adi krura
manusya ; (sukha 3, 9) | 2 vi. gramlna, gramya ; (sukha 2, 13) | “.
Indem ich die beiden zusatzlichen Bedeutungsangaben in die vor-
hergehende, iibrigens sehr geschickte Bedeutungsanordnung einbaue,
erhalte ich unter Verzicht auf die wohl iiberfliissigen SteUenbelege
folgende Ubersetzimg : ,,goJia-, m. Erster (Hauptling, Vorsteher)
eines Dorfes. einer Stadt, eines Districts ; Soldat, Soldner,
Panflurang Parab (Bombay 1904) ,, Jabdhah purusah " und im Subkommentar ,, gohe
purusah manuj^i/ah ; ebenso die funfte, von \asudev Laxman Shastri Pansikar
durchgesehene Ausgabe (Bombay 1922) „labdhah puru.fa}i‘‘\ t^gohe purusah
maviisyah Otto Bbhtlingk, Mrlvk'hakatika, d.L Das irdene Wagelchen, ein dem
Konig Pudraka zugcschriebenes 8ebaiispieb Ubersetzt (St. Petersburg 1877), p. 31 :
,, Den Kerl batten wir “ (in den ,, Anmerkungen “ p. 187 gar nichts) ; Ludwig Fritze
in seiner metrischen Ubertragung (Chemnitz 1879), p. 59 : Da haben wir den
Bursehen “ (ohne jede Bemerkung) ; Hermann Camillo Kellner, \ asantasena^
(Leipzig 1894), p. 52 : ,, Der Kerl ist erwisoht ! “ (ohne Anmerkung) ; Arthur William
Ryder, The little clay cart (— Haiward oriental series volume nine ; Cambridge,
Massachusetts, 1905), p. 31 : „ W'e've got our man (ohne jede Bemerkung) ;
Mrcchakatika (ed. Stenzler, Bonnae 1847) p. 31, Zeile 3 von oben (die Textstelle),
p. 191 (Glossierung der Prakrtabha$avyakhya : ,, lahdhah purusah ), p. 253
„ Adnotationes “ („ in farailiari Mahrattorum sermone usurpatur de
viro adolescente, homuncione eine Stelle, die sofort aus dem Folgenden versta^-
lich wird).
VOL. VITI. PARTS 2 AND 3.
858
W. WUST —
Kriegsmann, Sepoy ; Stationskommandant usw., uberhaupt ein stolzer
Herr ; Buhler, Liebhaber ; Bauer, bauriscb, Dorfbewohner, linkisch.
rob, ungebildet, ungescbliffen ; [Mann, Mensch, menschlicbes Wesen.
Person iiberbaupt] Da die in Klanimern stebende letzte Bedeutung
nur dem Zeugnis der Mrccbakatika zuJiebe angenommen ist und,
we icb oben p. 855 f. gezeigt babe, zugunsten einer sinnfaibgeren
tibersetzung aufgegeben werden muss, haben wir eine Bedeutungs-
summe, deren emzebie Teile unscbwer und vor allem iiberzeugend
aus der angenommenen Grund- und Ausgangsbedeutung ,, (in Aufputz
auftretender) Sprecber “ erklart werden konnen. Der Sachverhalt ist
so schlagend, dass icb micb mit wenigen kurzen, indischen und
ausserindiscben Hinweisen begnugen darf. Icb erinnere an die indiscbe
DorfVerfassung des Panchayat, die einen eigenen Vertreter des
Dorfes der Eegierung gegeniiber oder daneben den Dorfobersten, den
Lamhardar, vorsiebt. Bei Verbandlungen mit boberen Instanzen
musste dieser Dorfscbulze — und war es eine grossere Siedlung, Biirger-
meister gar oft als bauriscb, Imkiscb, rob, ungebildet erscbeinen.
Handelte es sich um die Verwaltung einer grosseren Gebietseinbeit,
so kam dafiir nur der Angehorige einer Adels- und Ksatriya-
Familie infrage, der zugleich, der ganzen Atmosphare jener friibmittel-
alterlichen Zeiten entsprechend, ein Kjiegsmann war. Die Gohils
als Clan waren solche Kriegsleute, und es scbeint mir kein Zufall, dass
in der oben p. 840 ' angefiilirten Liedstrophe die Wendung ,, fierce as
steel im Urtext ,, loha Icrura “ lautet. Von bier zur Definition der
Sukbabodhatika, der goha sei ,, Jcrura rnmnisya “ fiihrt eine gerade
Lime. Die Bedeutung ,, Buhle, Liebbaber “ ist, wie wir unten nocb
seben werden, aus einem Nebengebraucb der in *gobha- steckenden
\\ urzel voU zu versteben ; es ist das mannliebe Wesen, das abends oder
nacbts, kiuzum unter abenteuerlicheren Umstanden, durcb Gesang oder
^ ortrag die Gefuble der Angebeteten erregt ; daher aucb die von Stenzler
oben p. 85 / ' gebucbte Angabe, goho werde in der Umgangsspracbe
der Mahratten vom xnr adolescens, hmnuncio gebraucbt. So bleibt nur
nocb die Bedeutung „ Soldat, Soldner “ unbesprocben. Sie ist indessen
wie sonst. in KJammem. v> oj, .
ist nicht auf*
^ Ber Verfasser gibt keine Sanskritentsprechung
bemerkenswert ist ; keinerlei Beriehtigungen. Ein Wort gohil(l)a-
getuhrt. Die Abkurzungen bedeuten : pum = pumlinga ; [de] = de^I- oder deiva-
^ upadesapada ; mrccha = mrccbakatika ; sukha = sukbabodhatika
Hastalikhita) ; vi. = visayatySgopadesakulaka. Hastali-
anch nicht bei a' ^ ^'ofavala- nirgends etwaa ermitteln konnte —
m -d H. H. Wilson -nahm icb Prakritismus
an und ubersetzte demgemass versuehsweise.
ABrSCHE KULTUEGESCHICHTE
859
leicht zu erklaren. Denn entweder leitet man den Soldaten niederen
Standes aus den Verhaltnissen des Lehenssystems heraus von den
hohen und hoheren Dienstgraden ab oder man kniipft geradezu an
urtiimlicbe Verbaltnisse an, an den miles glorias us oder an den
altgermaniscben Krieger, der zu Beginn der Schlaclit den Barditus
ertonen Hess (Tacitus, Germania iii^), oder man zieht den Feld-
u’e{i)bel beran, dessen Nachglied ich unmittelbar < idg. *uoipl6s
(im Ablaut zu altindoariscb vipra-, m. „ Sanger, Dichter, Vorbeter,
Priester “) herleite Und damit sind wir bereits bei den hochst
lehrreichen und kulturgeschichtlich farbenreicben Verwicklungen
angelangt, die nicbt nur indoariscbe, sondern aucb idg. Verbal-
wurzeln mit der Bedeutung ,, sprecben “ gelegentlicb durcbmacben.
Der Mr Speaker ist der Prasident des engbscben Unterbauses ; die
Scbweizer Familie der Sprecher (z.B. von Bernegg) bat der Eidge-
nossenscbaft ausgezeicbnete Landvogte und Offiziere gestellt (z.B.
Fortunatus Sprecher im 17. Jabrbundert n. Cbr.), nbd. Palier, Polier,
eigentlicb ,,Sprecber“, bezeicbnet den ,, Obergesellen der Maurer".
In diese Bedeutungsrunde ist aucb das mittelindoariscbe goha-
Stiick fiir Stiick eingescblossen.
6. Vorstebendes semasiologiscbes Gesamtergebnis wird nacbge-
priift und vollauf bestatigt, indem icb jetzt die Untersucbung auf die
morpbologiscben Gemeinscbaftsverbande ausdebne. Icb kebre
damit gleicbzeitig zu dein altindoariscben Gohhila- zuriick, dessen
enge Beziebungen zu dem mittelindoariscben goha- von Cbarpentier
unter Hinweis auf das Yerbaltnis zwiscben paksa- und Paksila-
zwar bebauptet, aber nur unzulanglicb begriindet worden sind. In
der Tat kann und muss dieser Nacbweis in strengsinniger Folge-
ricbtigkeit und Scbliissigkeit erbracbt werden. Da es eine Moglicbkeit,
das Verfahren des Keim-Verbandes durcbzufuhren, so\’iel icb sebe,
nicbt gibt und die Priifung des Anklang-Verbandes scbon bei der
Widerlegung M. Bloomfields erscbopfend durcbgefiibrt •nmrde (oben
' Sehr auffallenrl ist deshalb der Bezug von bhata- zu * Y bhat ,, bha§arthe “ (bei
Westergaard, Radices und in der Ksiratarangini, ed. Liebich, nichts Wesentliches).
Vgl. auch die z.B. von Uhlenbeek fur „ onomatopoetisch “ erklarte bhatabhatiiy
„ einen glucksenden Laut von sich geben “. Femer waren fur die Beurteilung des
Xomens bhata- auch all seine Komposita heranzuziehen, wie z.B. ndbhata-, adj.
„ hervorragend, ausgezeichnet, ungewdhniich, heftig, leidenschaftlich “ mit auf-
fallender Bedeutungsverwandtschaft zu goha-. Schliesslich ware das Verhaltnis
zu bhatta-, m. ,, Herr, grosser Gelehrter, Doctor “ endgultig zu klaren. Man wurde
hiebei weit iiber Theodor Benfey, Abhandlungen der Konigl. Gesellschaft der
Wissenschaften zu Gottingen 2.3, Hi.st.-philol. Classe, Xo. 4, p 32-31. und liber andere
Gelehrtenmeinungen hinauskommen. Dies Problem sei hier nur gestellt !
860
w. viiisr —
p. 8-12 5), so bleiben noch folgende Aussagen iibrig ; Laut-
und Silbengrenze-Verband, der allgemeine EJang-Verband, Form-
Verband, Sippen-Verband. Der Laut- und Silbengrenze-Yerband
lasst erkennen, dass von den mbglichen Zerlegungen des Wortkorpers
Gobhila- rein hj^potbetisch, somit unbraucbbar sind ; G-ohhila - ;
Go-^bhila- (mit Avagraha me z.B. go^mbu-, gosnibhas- „ Kuhurin :
aber es gibt kein *abhila-) ; GobM-la- (es gibt kein *gobhi-) ; Gobhil-a-.
Die Auflbsung Go-hhila~ ist diejenige Bloomfields und somit widerlegt,
die Auflbsung Go- hh- tla- diejenige Charpentiers und somit glcichfalls
widerlegt. Folglicb bleibt nur nocb die Auffassung Gobh-ila- als
letzte iibrig und mbglich. Y ir wollen sie die W iist’scbe Analyse nennen.
Sie wird aucb durch den allgemeinen Klang-Verband bestatigt.
Die Bebspiele sind; *go]cila-, m. „ Keule, Pflug “ und gokida-, n.
,, Rinderheerde . Nun bestebt aber keine *\/ gucjk, die einer
Zerlegung goA-lla- bezw. *gok-ula- zugrundegelegt werden kbnnte,
vielmebr sind in beiden Y ortformen kila- bezrw. kula- als selbstandige,
nur m der Komposition sekundar gewordene Nomina ablbsbar. Also
kann Gobhila-, das mit goklla- und gokula- klanglicb-assoziativ
zusammenhiingt, nicht als Go-bhla- aufgefasst werden ; denn es
^bt, im Gegensatz zur vorherigen Prozedur, keine Mbglicbkeit, die
Komponente °-bhila- selbstandig aufzuzeigen. Folglicb bleibt, da
die anderen .inalysen entweder stark hypothetisch oder irrig sind,
nur die Zerlegung Gobh-ila- wiederum als die einzige iibrig. Anders
ausgedriickt .• wir .stelien vor der Notwendigkeit, im Falle Gohhila-
den FormA erband untersucben zu miissen, da vor aller wortkund-
ichen Namendeutung Aufklarung der morpbologiscben Ver-
unbedingtes Erfordernis ist. Oder, wa.s beim Beispiele
TO i a das Selbe besagt ; wir haben das -ila- Formans zu unter-
suchen. Zu diesem Zwecke babe ich ein umfangreiches Beleg-
material gesanimelt, riind 170 Beispiele Ibre eingebende Durcbsicht
80 und tibriges in der
419 " I^enfey, \oIlstandige Grammatik der Sanskritsprache
Sndne’r Al’t •’ f Renou, Grammaire Sanscrite ;
“manif U rr f = Macdonell, Vedic grammar;
iTa^da HiU n' ) P“Wished tfxt of the
in die .Sansknffo™”'^’ Personennamen (buddhistische Namen sind
indischen S Theophil Gubler, Die Patronvmica im Ait-
Spiegel de^KuW 1 I ’’ Indogermanische Eigennamen al.s
® ^^^turgebcluehte {Heidelberfr ' iViiof c r /•
Gegensatz zu den vorstehenden r " ’ ■ e"
Bildungen, die aus der T?p l f ii *■ "'cht eigens bezeichnet). Kiare
^ttadlieh ausgeschiedtn. selbstver-
AEISCHE KULTUKGESCmCHTE
861
zeigt, dass die Bildungsweise vom Vedischen ins Sanskrit hinein
klarlich zunimmt dass ganz verschiedene voraltindoarische sprach-
geschichtliche Ausgangsstufen anzunehmen siad dass diese aber
durchaus nicbt einheitlich und insgesamt fiir die Beurteilung von
Gobhila- herangezogen werden miissen, sondern dass die Sprach-
vergleichung bereits aus diesen Mbgliclikeiten und Vorstufen eine
bezeicbnet, die allein auf das Paar mittelindoarisch goha- : altindo-
arisch Gobhila- zutrifft. Icb umschreibe sie init den Worten Brugmanns.
Grundriss der vergleichenden Grammatik der Idg. Spracben 2, P,
p. 360: ,, die ganze Kategorie der ai. Deminutiva wie ai. vrsald-s
‘ Manniein ’ und der ai. Eigennamenformen wie ai. Deiila-s [ist]
mit Sicberheit den uridg. -lo- Formationen zuzuweisen ‘■®. Angesicbts
der gesamten klaren Sacblage des Eigennamens Gobhila- und seiner
Fortsetzer bandelt es sicb also unmissverstandlicb um sekundares
-{i)la-, von urspriinglicb bypokoristischer, vohlgemerkt nicbt
deminutiviscber Funktion Icb deute die Gescbicbte des so
bescbriebenen Formans mit einigen wesentlicben Stricben, obne
VoUstandigkeit beabsicbtigen zu wollen, an. Einwandfrei ist es im
Altgriecbiscben nacbzuweisen ; ich nenne aus den bieriiber vorlie-
genden Sammlungen ein paar Beispiele ’ : ’ApKiXos, OepaLXos,
&vp.l\o£, ’OpyiXo'S, ’ OvdalXos, ttolkIXos, HiotXos, Ta^lXos,
XoiplXos. Aussergewdbnlicb lebrreicb ist Eo^iXos, das natiirlicb
zu (jo(f>6s gehort. Wer es entgegen alien Einsicbten und
* Der Atharraveda z.B. hat iiur 2 (3 ?) Belege ; dies zur Erdrterung p. 841 f.
^ 1. Vorhergehende NuUstufe des Wurzelelements ; 2, vorhergehende \olIstufe
de.s Wurzelelements ; 3. < *-3lo- (anila-, jalija- :jata-) ; 4. < *-iro- ; 5. < *-i-ljro-
usw. Vgl. noch Renou, Granimaire .Sanserite 1, p. 218 unten : „ ira- da- issu sans
doute de i + ra- la- (ef. anila- : aniti . . .) figure dans quelques formes sans nettete".
Ahnlich meint Whitnev, Grammar®, § 1189 bei Besprechung de.s primaren -la-:
,, Manj' words ending in la are of obscure etymology “.
® Zusammen mit Anmerkung 1 beweist dies erneut, dass Albrecht eber Gobkda-
falsch beurteilt hat.
* So zuletzt Ernst Eraenkel, .4rtikel Xamenwesen in Pauly-Wissowas Real-
Encyclopadie der classischen .Alterturoswissonschiift. Neue Bearbeitung 16, Sp. 1611
unten-1670 unten, namentlich al>er Sp. 1637 f. Auch diescr .Autor unterstreicht das
Hervortreten des hypokoristischen Grundzugcs (Affekt, Anteilnahme, Herkunft.
(Gn)lust, Zugehorigkeit) und das Zurucktreten des deminuierenden, der erst ein-
zeLsprachlich reichlicher zu belegen ist. In Devila-, Dattila- : Devadaita- erkennt auch
Fr. idg. hypokoristisches -{i]la- (a.a.O. Sp, 1637). — Die hy'pokori.stische Funktion
des -ila- wird bereits von Panini ausdriicklich gelehrt. v, 3, 79 und 83 (vgl. Hilka
a.a.O. p, 56 oben und unten). Aus Hilka p. .>8ergdjt sich die ja ganz gelaufige Tatsaehe,
da.ss der mehrtciligc Vollname. der beim Kurz- oder Kosenamen Gobkda- <c *gobhn-
Pate gestanclen ist, nie zu ermittein sein wird. Fur die Deutung von Gobhila- selbst
1 st das unerheblich.
® Brugmann a.a.O. p, 368 unten, 376 : Solmsen-Fraenkel a.a.O. p. 130 u.. 133.
862
W. WUST —
Gegebeateitea zu (jJt'Ao? stellen und So-(f>lXo 5 analysiere:
wollte, der wlirde so liandeln wie Bloomfield mit seiner irrigen(ro-6^;/"
Analyse. EorfiiXos lehrt an seinem Ende innerhalb des allgenieiii
idg. Bereiches, dass die ZerJegimg Gobh-ila- tatsachlich auch a priori
viel W ahrscheinliclikeit fiir sich hat. ~ Aus dem Litauischen ist
erzilas ,, Hengst “ hieher zu stellen. Die altgriechische Gruppe hat
Sophus Bugge, K.Z. 32, p. 78 mit dem armenischen Sekundarsuffix
-il verkniipft und als Beispiele namhaft gemacht ; Jcat'il ,, Tropfen.
Trdpfchen rmCil „ kleine Laus orpl „ Laus “ Im AJt-Iranischen
dagegen sind keine sicheren Belege feststellbar eine Tatsache, die
gar nicht ernst genug genommen werden kann. Tut sie dock mit
ziemlich hoher Wahrscheinlichkeit dar, dass Gobhila- kein iranisches
Lehnnort seiu kann. Die Linie des entwicklungsgeschichtbchen
V erlaufs fiir altindoarisch -ila- ist oben p. 860 f. schon skizziert worden.
Es ist nun klar, dass es nicht nur unnotig weit fuhren wurde, alle -ila-
Belege des oben beschriebenen Typus in vollem Umfange hier duich-
zusprechen, sondern dass es auch dringend geboten scheint, von so
Die gotisch-germanischen Eigennamenformen wie Oflila, Tuldila, Wuljila,
ahd. II igilo usw. usw. gehdren als urspriingliche -g/on- Stamme mit ihrem doppel-
deutigen -i- nur mittelbar hieher.
^ Bartholomae, Altiranisehes Worterbuch Sp. 1947 verzeichnet nichts, unter
■tra- nur nicht Hiehergehoriges, wie z.B. vl-ii-ra-, adj. Auch bei A.V. Williams Jackson,
Ave.,ta grammar p. 213 f,, 228 f. ist nichts zu ermitteln, desgleichen nicht bei Justi.
Iranisches Xamenbuch p. 521-^ „ Verzei.hniss der Namen nach den Ableitungs-
xen sowie der gekurzten und mit Kose-Affixen versehenen Namen (entbalt
unter ,, el, il (armen.) al (neupers.) “ nur Material, das entweder gar nicht hier-
hergehort oder fragwurdig ist, wie z.B. skythisch Sagillus). — Im Mantrabrahmana
^ (lie Forme! vor ; „ ahura idarh te paridaddynyafi^y-my"
was araporewala a.a.O. p. 143 ubersetzt : ,, O Ahura, here to thee I deliver so-and-
"" ii*'- n,ber ganz unmdglichen Schlussfolgerungen benutzt.
ura so eine ,, Aryan deity sein, zugehbrig einer ,, period before the tw'O peoples
sepMated. Ahura needs no comment “ (a.a.O. p. 146 unten). Dabei kdnnte die
o orm gar noch nicht einmal mehr der arischen Gemeinschaftsepoche angehoren.
wei sie en gemeinirani.schen Lautwandel ~s- > durchgeraacht hatte. Bestenfalls
kannessich also nur um ein rein iranisches Lehnwort innerhalb des Alt-Indoarischen
an e n, «o ur ich, ausscr Bekanntem, bei Gelegenheit noch das eine oder andere
ei.spie iznbringen hoffe. Es kann also gar keine Rede davon sein, dass hier ein
L berrest vorliege „ of a very ancient tradition going back to the period when the
ndo-lranians were living together as a united race “ (a.a.O. p. 147 oben). Ausserdem
musste die Wortform Ahura zunaehst einmal sorgfaltig anhand der beigebraebten
ananten {ahur, abhura, antara, ahuri, abhuri ; Knauerubersetzt „ Feuer des Magens ‘‘I
geprutt werden. So bleibt eine unbestreitbar lehrreiche, aber durchaus verwickelte
rage, die, zugunsten der gereinigten iranischen Lehnthese beantwortet, vieUeiebt
det '^'^'•‘‘odnis des Eigennamens GohhiZa- (: \/ g(a)ub
Gobhilaurhv^*^ einmal wesentlich werden konnte. Denn Mantrabrahmana und
plan “ (Oldenl!! w-orks have been composed together and on one common
towenberg a.a.O. bei Taraporewala).
AEISCHE KULTURGESCHICHTE
863
allgemeinen Zuordnungen, wie sie beispielsweise Cliarpentier (oben
p. 846) ausgesprocben bat, loszukommen und dichtere, strengere
Bedingungen des von mir so genannten Formenkatalogs zu finden.
In diesem Sinne babe icb den gesammelten Stoif nocb einmal durcb-
priift und zwei Gruppen gebildet, die ibrem Form- und Edang-
Verband nacb am engsten zu GobJiiU- geboren. Wer Gohhila- als
reine Wortform auf sicb virken lasst, wird neben der Dreisilbigkeit
vor allem die Vokalfolge -o-i{l)a' bemerkenswert finden und nacb ibr
den Einbau der zum Stammgefuge gehorenden Media aspirata in
der ilitte des Wortkorpers. Damit ist der Formenkatalog bescbrieben,
und alles bandelt sicb nun darum, genugend Beispiele fur beide
Sondergruppen zu finden. In die erste Gruppe, der wir infolge des
bekannten altindoariscben, aufs Idg. zuruckgebenden Vokal-
parallelismus unbedenkUcb auch die -e-ila- Belege zurechnen diirfen,
geboren binein ^ :
Omila-, m.n.pr. ; oma-, m. ; kokila-, m. ; koka-, m. (nebst
kaukila- und kaukili-) ; Kosthila-, m.n.pr. : kostha-, m. ; Kausila-,
m.n.pr. (; Kausika-, m.n.pr. und mittelbar : kosa-, m.) ; \gopila-,
adj. : gopd-, m. (das scbon firubzeitig als Worteinbeit empfunden
vorden ist) ^ ; \ghosila-^ m. : ghosa-, m. ; johila- = griecb. ZcotXos
(das altindoariscbe Wort ist sebr lebrreicb, weil es als reines
Lebnwort die Starke des indoariscben Klang-Verbands zeigt) ;
Totild', f.n.pr. : totddri-, m.n.pr. < tota + adri - ; Bodhila-,
m.n.pr. : bodhd- oder zu bodhi-, m. (so Hilka, p. 69) ? ; Samila-,
m.n.pr. : soma-, m. ; Horila-°, m.n.pr. : hard- ? Man bemerke
das Vorwiegen der n.pr.-Bildung !
^chekila-, adj. : cheka-, adj. ; ]devila-, m.n.pr. : devd-, m. ;
jdiemla-, m. und adj. : fhena-, m. ; Rebhila-, m.n.pr. : rebkd-
adj., m. und m.n.pr.
Form- und Klang-Verband zwingen zu dem Scbluss : Gohhila-,
m.n.pr. : *gobha-. Die vorlaufigen Darlegungen p. 838 f., 847^
unten. 852-, 861-3, auf die icb insgesamt nocbeinmal nacbdruckbcb
verweise, sind durcli diese Scblussfolgerung zugleicb gerecbtfertigt
^ Ich habe anhand von Whitneys Roots, anhand von pv. und Schmidts
Xachtragen samtliche Wurzeln des Typus (k^)iju m^h gepriift ; es ergab sicb kem
weiteres -Ua- Material mehr. i-' bedeutet irgendeinen Konsonanten. m Media.
Kokokila- entspricht dem Formenkatalog infolge seiner ViersiJbigkeit nicht, ebenso
auch fkarketila- nicht.
“ Infoigedessen darf man sich auch von vorneherein nicht von der Analyse go-pd-
leiten lassen, da diese fiir *< 70 - bha- unfruchtbar ware, wie oben p. 847* f. bewiesen
Worden ist.
864
W. WUST—
und abgerundet. Aber die Ergiebigkeit dieser Schlussfolgerung
erschbpft sich damit noch keineswegs. Wir haben Gobhila-, m.n.pr.
dem anderen Eigennamen Rebhila- dichtest gegeniiber gestellt und
Rebhila- mit rebhd- zusammengeriickt. Dieses Adj., m. und m.n.pr.
rebhd- aber ist eine klare, unzweifelhafte Xominalbildung der rihh.
die scbon im Rgveda bezeugt ist und so\’ieI wie ,, lobsingen, rauschen.
singen “ bedeutet. Diese Formenlage fiihrt auf die einleuchtende
Proportion :
rebhd-: ^/ ribh = *gobha- (+ mittelindoarisch goha-) : j'-
Es ist augenscbeinlich, dass anstelle dieses x *\/ gitbh eingesetzt
werden muss, genau so wie wir — und nun bilft uns der Reim-
\ erband dieses *gobha- weiter ! — zu ksobha-, m. V Icsubh, zu lobha-.
m. lubh, zu sobha-, m. -\/ subh, zu stobJia- m. stubh tatsachlich
vorfinden Und nicbt nur dies ! Die Wortformen jSoMd- und
(man bemerke das -ka - !) erscheinen auch als Eigennamen (Hilka
a.a.O. p. 125), und als VoIIname gar Rsyalobha- (Hilka a.a.O. p. lU
unten). Eine treifendere, iiberzeugendere Bestatigung fiir das inner-
balb des Altindoarischen erschlossene *gobka- lasst sich, in seinem
^ erhaltnis namentlich zu unserem Gobhila-, schlechterdings nicht
vorstellen. Und diese Bestatigung wird ihrerseits noch dadurch unter-
strichen, dass Charpentier oben p. 847 fiir sein Postulat *g’f6uhho-
keinerlei Parallele im Idg. hat nachweisen konnen, wahrend lobha-
beispielsweiseuntadeligeGleichungmit altengl. leq/’,,Erlaubnis“ bildet.
Die zweite Gruppe, in der Lautungen mit -m^h-ila- Ausgang
zusammenstehen, vervollstandigt das gewonnene Bild. Ich beschranke
mich, aus Raumnot, in diesem Zu.sammenhang, darauf, aus dem
ge.samten Stolf die Belege bier zu nennen, welche der scharfsten,
einschrankendsten Bedingung geniigen, namlich als Media aspirata
inlautend -bh- zeigen. Das sind ;
^kutnbhila-, m. : kunibhd-, m. ; \nabhila-, adj. (das zu ndbki-.
nif. gestellt wird, aber auch zu nabha-, m. gehoren kann) ; Nikuoi-
bhld-, mf. n.pr. : mikmnbha-, m. ; Rebhila-, m.n.pr. ; rebhd-.
adj., m. und m.n.pr.
Es ist kein Zufall, dass Rebhila- in beiden Sondergruppen vertreten
ist , es ist kein Zufall, dass Rebhila- und goha- in der Mrcchakatika
sich begegnen, sowenig wie es Zufall ist, dass, wie ich noch zeigen
f A selbstverstandlich niemandem einfallen, zu analysieren kso-bha-,
o » , so-bha-, Mo-bha-, und in diesen Bildungen je zwei verschiedene Worte entdecken
* Diese Pberlegung allein zeigt schon, wie toricht und zufallsbedingt es
gobha- in go-hha- zu zerlegen.
ARISCHE KULTURGESCHICHTE
865
werde, rebM- und gu im IRgveda eng beieinander stehen. Rebhila-
und Gobhila- gehoren als untereinander nachstverwandte Bildungen
in den gleicben Klang- und Form- wie Sinn- und Bedeutungs-
Verband, imd es ist ungemein lehrreich, an den wenigen Stellen, wo
Rebhila- als Nomen 'ftroprium innerhalb des altindoariscben Scbrifttums
auffcritt, nocb den Nachball seiner alien appellativischen Grund-
bedeutung zu vernehmen, so Mrcchakatika p. 43, Zeile 14 (ed.
Stenzler) ; ,,aho sadhu bhavarebhilakena gitam“, so Mrcchakatika
p. 44, Zeile 6 ; ,, vayasya susthu khalv adya gitam bhdvarebhilena “
7. Die Beweiskette hat sich nahezu geschlossen. Nur eine kleine
Liicke klafft noch in ihr, die Priifung des Sippen-Verbandes, die,
von einer bemerkenswerten Kleinigkeit abgesehen, ganz normal
verlaufen karm Von Gobhila- sind folgende Nomina regelrecht
abgeleitet ; gobhiliya-, adj. ,, zu Gobhila in Beziehung stehend
(pw., ahnlich das PW. ; keinerlei Nachtrage, Verbesserungen und
Zusatze, nichts bei Schmidt, Nachtrage) ; fgobhihka-, \gaMilikd-,
beide Patronymika ,,nach der Lehre der Grammatiker ‘‘ (Bohtlingk.
Panini ’s Grammatik, Leipzig 1887, p. 320* sowie Gubler a.a.O. p. 51) ;
schliesslich ,, Gaubhila n. Gobhila’s Grhyasutra Grhyas. 2, 94 “ (so
Schmidt, Nachtrage s.v., nichts in PW. und pw.). Das letzte \rddhi-
Derivat zum Eigennamen Gobhila- ist allerdings kostbar genug.
Es zeigt nicht nur, weil es im strengsinnigen Form- und Klang- Verband
mit anderen Patronymicis wie Kaukila-, Kausila-, Sautmla- zusammen-
steht, dass die bei Monier- Williams - gebuchte Vanetas lectionis
Gobida- sinnlos ist, sondern es veranschaulicht auch noch einmal
schlagartig, dass die umfassend geforderte Zerlegung Gobh-ila- die
einzig mogliche und die einzig richtige ist.
8. Sind die bisherigen Ergebnisse, und zwar sowohl die, welche
auf Widerlegimg alterer und j lingerer Deutungsversuche des Eigen-
namens Gobhila- abzielten, als auch die, welche im Verfolg einer neuen
Betrachtungsweise gewonnen wurden, richtig erarbeitet, dann bietet
sich uns, sozusagen von selbst, jetzt am Schlusse unserer Uberlegungen
eine ausserindoarische Ankniipfung als wechselseitige. wertvollste
Nachpriifung und Bestatigung ; die altiranische ( = altpersische)
V g(a)ub ,, dicere “. ^/ g(a)ub ist erst vor einigen Jahren noch von
F. B. J. Kuiper, Acta Orientalia 12, p. 268 unten als .. bisher unerkliirt
bezeichnet worden, eine Bemerkung, die natiirhch nur den ausser-
iranischen Befund, nicht aber den inneriranischen Belegstand
^ Ganzlicb verblasst Mrcchakatika p. 67, Zeile 10.
* Bemerkenswerte Zusammensetzungen sind mir nicht bekannt geworden.
866
W. Wt'ST —
kennzeichnen vrollte aber wissenschaftsgeschichtlich jedenfalls irrig
ist. weil schon lange vorher Meillet, wie war gleich sehen werden.
eine durchaus annehmbare Deutung gegeben hatte. Denn in der
Tat gebort die ■\/g{a)uh, ahnlich wie das von mir jiingst untersuchte
altiraniscbe casman-, n. Auge zu den iranischen Worten, die
dem Gesamt -Iranischen, zeitlich und vielleicht auch mundartlich
gesehen, eignen. Innerhalb des Alt-Iranischen findefc sich ■\/g(a)uh
bemerkenswerterweise nui im Altpersischen, und da wiederum nur
in der Bahistan-Inschrift in einer Gebrauchssphare, die verbliiffend
mit derjenigen des mittelindoar. Nomens goha- iibereinstunnit. Die
\'gaub wird namlich im Altpers. grosstenteils vom sich verstellenden
Sprechen niilitarischer Wesenheiten angewendet (sechsmal in Bezug
auf eine rebellierende Einzelperson, sechsmal in Bezug auf ein
rebellierendes Heer, einmal in Bezug auf eine rebellierende Provinz ;
demgegeniiber nur einmal im Zusammenhang mit der richtigen
Aussage eines Darius-Heeres). Dies kann kein Zufall sein. Die
militarische Atmosphare der Rebellion ist, selbst wenn wir die
literarischen Voraussetzungen, ahnlich wie beim Text der gotischen
Bibel, bedenken, geradezu iiberwaltigend, auch rein zahlenmassig.
Darnach erscheint \/g{c()ub mit einem reichen Formenstand im Mittel-
Iranischen aus dera es, wiederum mit mannigfaltiger Entwickelung.
sich ins Neu-Iranische hinein niehr oder minder deutlich entfaltet, nicht
ohne auch noch im Armenischen einige Lehnspuren zu hinterlassen®.
^ ^ gi* z.B, die Formen des ind, und conj. praes. med,, gaubataiy und gaubatciig
Pas ganze Material ist bequem zu uberbiicken bei Bartholomae, Altiranisches Worter-
buch »Sp. 482 f. {insgesamt 14 StelJen ; keinerJei ,, Nachtrage und Verbesserungen ‘ »
auch nicht im ,, Beiheft wozu man noch Tolman, Ancient Persian Lexicon p- 8”
unten hinzunehme (Ansatz gub !). Roland G. Kent, The recently published Old Persian
inscriptions [Reprinted from JowrnaZ of the American Oriental Society, vol. 51, number 3,
pages 189-240] enthalt leider keine neuen Funde zu g{a)ub, wie „ Concordance and
glossaiy “ p, 236 unten ergebeo.
^gl. zuletzt H. S. Nyberg, Hilfsbuch deg Fehlevi II. Glossar (Uppsala 1931),
P- 82 : „ go^isn dag Sprechen, das Reden ; Wort . . . — zu guftan “ ; p. 84 f.
»» guftan go^ . . . sagen . , . guftan scheint einheitlicher Ausdruck fur ‘ aussem ’ zu
sem . . . Pers., Texte zum mazdayasniscfaen Kalender (— Uppsala Universitets
Arsskrift 1934. Program 2), p. 76 oben, wodurch W. Henning, Zeitschrift fur Indologie
und Iranistik 9, p. 184 (in seiner Poktorschrift ,, Das Verbum des Mittelpersischen
der Turfanfragmente ‘ ) erganzt wird. Pie Belege aus dem Soghdischen finden sich
bei Robert Gauthiot und E, Benveniste, Essai de grammaire Sogdienne 1, p. 129,
145 , p. 12 (vgl. auch 2, p, 220, Sp. 1 unten und f.). Ich nenne daraus an Formen ;
yofi- ,, louer, vanter ; yaipt, *y6^ot, 3. pers. sing. ind. praes. act. ;
parte.
® Genannt seien : vomeupers. (in diesem FalJe = Pahlavi Vendidad, vgl. Horn,
silch ^uper&ischen Etjmologie p. 263) aguftdr „ nicht sprechend “ : neuper-
mar „ Rede neupereisch guftan „ sprechen “, das bei Horn a.a.O. No, 926
ABISCHE KtTLTURGESCHICHTE
867
Wesentlich fiir die wissenschaftliche Beurteilung dieses sprachJichen
Gesamtvorgangs aber ist, soviel ich sehe, bisher imrner gewesen, dass
guh, gauh als dialektologiscbes Merkmal gegolten hat, eine Ansicht,
die sich zum erstenmal wohl bei H. Hiibschmann, Persische Studien
p. 94 (zu Horn, Grundriss der neupersischen Etymologie No. 926)
deutlich ausgesprochen findet : ,, Wrzl. gub ‘ sprechen scheint
spezifisch persisch (der alten Provinz Parsa entstammend) zu sein
fiir vac der andern Dialekte. Im Kurdischen wird nur das Ptcp. perf.
von der AVurzel guh gebildet und aucb dies konnte entlehnt sein”
(vgl. unten Anm. 1 !) Diese Communis opinio wird im Lichte des
neuen Zusammenhangs mit altindoarisch Gobhila- genau so sicher
einer durchgreifenden Umformulierung und tlberlegung der
dialektologischen Probleme bediirfen^, wie schon jetzt der bis heute
doppeldeutige Auslaut der altpersiscben \/ gub, gaub sicher -bh
gewesen ist. Das altindoarische n.pr. Gobhila- hat diese Entscheidung
herbeigefiihrt
AVie steht es nun aber auf der altindoarischen Seite mit der
als Stichwortartikel behandelt ist (die Pamir- Dialekt-Belege Homs, namlich „ wa^.
joy-am, sar. “ finden sich im „ Grundriss der Iranischen Philologie “ nicht
wiederholt) ; sonstige neupersische Formen bei Horn, Grundriss der Iran. PhUol.
1,2, p. 47, rein mundartliche Fonnen des Xeu-Iranischen im ,, Grundriss “ 1, 2, p. 353,
354, 363, 372, 388. Soein, Grundriss der Iran. Philol. 1, 2, p. 261 bespricht knrdisch
gofeii. (raundartlich gohtin) ,, sagen “, die man bei Mann-Hadank weiter verfolgen
mag. Armenisches Lehnwort ist jata-gov ,, Fursprecher “ (Horn, Grundriss 1, 2,
p. 47, 63 unten und f.). Strittiges bei C. Salemann, Grundriss 1, 1, p. 270. Schon
Aug. Friedr. Pott, 'Wurzel-Worterbuch der Indogermanischen Sprachen 5 (Detmold
1873), p. 255 gibt einen ganz hubschen Querschnitt durch die inneriranische Geschichte
der Wurzel gub, gaub, indem er gleichzeitig auf Spiegel und Lerch verweist.
' „ Altp. gaub nur im Medium vorkommend = ' sich nennen ’ im Unterschied
von Bah ‘ sprechen, sagen, befehlen’, pass. ‘ heissen’ (skr. ra.s)“. Ahnlich zu Obigem,
a.a.O. p. 116^ mit wichtigem Schrifttum ; femer Wilh. Geiger, Grundriss der Iranischen
Philologie 1, 2, p. 414 unten und Wolfgang Lentz, Die nordirauischen Elemente in
der neupersischen Literatursprache (= Zeitschrift fur Indologie und Iranistik 4,
p. 251-316), p. 305 ; „ der Gebrauch der Wz. giifidn, die sonst nur im Sogd. vorkommt
(AXDREAS)[,] ist ein hervorstechendes Merkmal des Persischen gegeniiber den
nbrdlichen und zentralen Dialekten *■.
^ Da die Vervvendungsueise der y/gaub „ sich falschlich ausgeben als jemanden “
gegenuber der etwa der y/ Bah ganz deutlich verschieden ist, ist uberhaupt zu envagen,
ob die ursprunglich angenommene mundartliche Spaltung nicht in Wirklichkeit der
bekaimten ahurisch-daevischen Glaubens- und .Sprachspaltung zuzuschreiben ist.
Fur die altpers. Inschriften und ihr religionsgeschichthches Verhaltnis zum Avesta
ware die Bejahung dieser Frage von weittragender Wichtigkeit.
’ t ermuten konnte man ja dieses auslautende -bh bisher schon mit einiger
Sicherheit und Wahrscheinlichkeit, da die altiranischen, auf-6 ausgehenden Wurzeln
durchwegs ein -b < -*bh haben, so z.B. grab, ^dab, zamb usw. ,. Bei der Seltenheit
von idg. b gehen die meisten b des Iranischen usw. auf bh zuriick ", bemerkt treEFend
Eduard Schwyzer, Griechische Grammatik I, p. 297*.
868
\v. wtrsT —
Entsprechuiig der altpersischen y/ guh, gaub, mit der urzel *gubk ?
Auf diese Frage, die schon oben p. 864 in anderem Rabmen gestreift
worden ist, lasst sich fiirs erste, wie es scheint. nicbt \’iel Sicberes
antworten. Das Dasein einer * gubh ist in keiner der urzellisten.
also weder bei Delius, noch bei Westergaard, noch bei hitney, noch
in der Kslrataranginl (ed. Liebich) nacbweisbar, und aucb in meinen
eigenen grossen lexikograpbiscben Sammlungen babe icb nicbts
derartiges aufgespilrt. Und trotzdem darf eine *y/ giibh angenonuneii
werden. Sie zeigt sich, ahnlich wie bei cbemiscb-pbysikaliscben
Vorgangen, in sogenannten Reaktionen, die wir nur auftnerksam zu
analysieren brauchen, um ibrer unsicbtbar-sichtbaren Wirksamkeit
beizukommen. Unsere * y/ gubh wird erstens bewiesen durch ihre
nominalen Absenker Gobhila- und goha- (< *gobha-). Sie wird
zweitens bewiesen durch das eigentiimlicbe Verhalten der y/ gub
jjVerhiillen. verbergen die laut Ausweis der Whitney schen Roots
keine einzige, nicbt nominale und nicbt verbale, Form mit -bh-
gebildet hat obwohl dies, nach alien sprachlichen Vorgangen, die
wir kennen, durchaus nicbt femgelegen batte. Mit anderen M orten.
die y/ guh ist der \/ gubh ausgewichen, ein Verhalten, das wir nach
dem richtungweisenden Aufsatz A. Meillets ,, Sur les eifets de
I'homonjTnie dans les anciennes langues indo-europeennes
(= Cinquantenaire de I'ficole Pratique des Hautes fitudes, Paris
1921, p. 169-80) zu wiirdigen wissen^. *y/ gubh wird drittens wahr-
scheinlich gemacht durch das schlagend proportionale Verhaltnis,
das zwischen ihr und y/ gii einerseits, sowie zwiscben y/ stu und
y/ stubh andererseits besteht. Denn stubh ist bereits vedisch, und
umgekehrt fehlt sie ini Iranischen. Man beachte gebiibrend auch
neben dem Reim-Verband den Bedeutungs-Verband, der y' gubh,
etwa ,, aussern, sagen, sprechen mit y/ stubh ,,jucbzen, trallern
verbindet. Auf den weiteren Reim- und Form-Verband, der
hiniiberfiihrt zu den y/y/ ksubh, lubh, subh, ist bereits oben
aufmerksam gemacht worden. Die Beweismittel sind noch nicht
erschopft. Das Dasein einer y/ gubh kann sprachpsychologisch auch
durch die Proportion umschrieben werden ;
v' tuTij ; y/ tubh ^ = r ; y/ guhj. Als r muss *y/ gubh eingesetzt
^ Hanns Ocrtel, Roots and verb-forms from the unpublished parts of the
Jaiminiya Brahmana (= Journal of Vedic studies 1, p. 120-68) enthalt weder ova
zu 's/ guh noch irgend cine gubh.
^ Bamit sind meine Darlegungen oben p. 840 und p. 849 unten abgeschlossen.
^ Gerade die y' iubh veranschauiicht, dass literarisches Nicht -Belegtsein und
erbverwandtsthaftlifhe Beziehungcn sich gar nicht auszuschliessen brauchen.
ARISCHE KULTURGESCHICHTE
869
werden. Schliesslich hat A. Meillet mit dem ihni eigenen Spiirsinn
die grossen idg. Zusammenhange aufgedeckt, in die sich auch
*\/ gubh einreiht, wenn er, 3ISLP. 11, p. 183 f. ausfiihrt : ,, en
\deux perse, en lituanien et en vieux prussien, I’addition d’un
elargissement b donne a des racines signifiant ‘ faire un bruit, crier,
chanter ’ le sens de ‘ dire, parler ’ Aus den Beispielen a.a.O.
p. 183 m. ; ,, skr. joguve ' il fait entendre un bruit ', gr. v. si.
goi'orii ‘ bruit etc., et avec elargissement d, lit. gaudziaii, gausti
‘ tbnen ’ ; v. pers. gaubataiy, persan giiftari ‘ dire ’ Meillet legt
weiter dar ; ,,L’iranien^ et le baltique, les seuls dialectes indo-
europeens ou apparaisse le fait, ne permettent pas de decider s'il agit
ici de i.-e. b on bh ; mais b est invraisemblable a priori et d'ailleurs-
on a bli dans arm. olb ‘ gemissement ’ en regard de gr. 6XoXvt,<u
. . Es erfiillt mit Genugtuung, diese feinsinnigen Ausfiihrungen des
ausgezeichneten Gelehrten durch zwei Feststellungen zu erganzen.
einmal dass auch das Indoarische der oben bezeichneten Gruppe
zugehbrt haben muss — wie die iibersehene V stubh und vor allem
der Eigenname Gobhila- zeigen — und dass durch Gobhila- insbesondere
die von Meillet schon richtig entschiedene Frage, ob -b oder -bh,
unniissverstandlich zugunsten von -bk nochmals beantwortet wird.
Nach dem Yorgetragenen kann ich mir durchaus denken, dass
selbst hartgesottene Zweifler an eine gubh zu glauben geneigt sein
werden. Wer aber auch angesichts dieser Nachweise noch nicht
iiberzeugt ist, der gehe mit mir noch einmal vom Gesamt- Grund-
sprachlichen zuriick zum einzelsprachlich Alt-Indoarischen, zur
Ausgangswurzel der *\/gubh, zur gu-. Wir wollen noch eine
Bedingung setzen und zusehen, ob und wie sie sich erfiillen lasst, die
Bedingung namlich : gehoren gu und gubh wirklich zusammen.
dann muss sich dies nicht nur im Theoretisch-Konstruktiven, sondern
auch im Lebendigen einer Sprache, in Wort und Schrifttum zeigen.
Und dies lasst sich in der Tat so eindringlich zeigen, dass wir Meillet
sein Stichwort gu dankbar zuriickgeben kbnnen. Ich halte folgende
Tatsachen fiir beweiskraftig. 's/ gu, die so\'iel wie ,, ertbnen lassen,
laut aussprechen, verkiinden “ bedeutet (mit prati ,, horen lassen “),
Denn tubh ,, (lurch einen .Schlar: verletzen “ gehort mit griechisch
,, niis.shan(Jle, stosse “ zu.sammen (I. Schcftelowitz, Zeitschrift fur Indologie und
Irani.stik 2, p. 279 unten). Andcre Iiterarisch nicht belegte Wurzeln, denen im
Bildetypus sich *-,/ gubh anreihen konnte, sind 1 / sku(m)bh (Westergaard p. 222)
und y/ stumbh (Westergaard p. 22.3).
* Man Tergleiche auch noch Soghd. (r'/S *wag „ parler, sprechen “.
■ Bei Oertel a.a.O. hierubcr nichts.
870
W. WTIST
erscheint im altesten indoarischen Text gedanklich eng\'erbuiiden
mit rebhd-, dem Ausgangswort 2 ru Rebhila-, dessen nahe "Verwandt-
schaffc mit Gobhila- wir oben p. 864 kennen gelernt haben. Die Stelle
ist Rgveda i 127, lO-i-g ;
„prdti ydd tm havisman
vlsvdsu hsdsu joguve
dgre rehho nd jarata rsundm
jimir Mta rsunam"\ in Geldners Ubertragung; ,, Wenn ihn
an alien Orten der Opfernde anruft, lasst der Hotr wie ein Barde den
Weckruf ertonen noch vor den Morgenstrahlen, der glubende Hotr
(vor) den Morgenstrahlen “ (Oldenberg, Noten, ohne Bedeutung).
Wir notieren die ausschliesslich priesterliche Verwendung, die auch
Rgv. i 61, 14'^'* fiir s/ gu gewahrleistet ist, und sehen uns die zweite
Stelle an. Es ist Rgveda v 64 2'^'^ :
,, sevaiii hi jdrydm vdm
visvdsu ksdsu joguve
Ich iibersetze, indem ich joguve mit Oldenberg, Noten 1, p- 357
als 3. Sing, fasse und auf den Stabreim jdrydm joguve aufmerksani
mache ; ,, ein holdes Buhlenlied namlich fiir Euch beide singt er
laut an alien Orten “. Die Verwendungsweise ist zwar auch noch
weiterhin priesterlich, aber die Wendung ,, ein holdes Buhlenlied
fiihrt uns ins tagliche Leben hinein und in gerader Linie zu der
Bedeutung ,,jdra, upapali''", die Hargovind Das Sheth dem Nomen
goha- zuerkannt hatte. Ein Kommentar ist wohl iiberfliissig.
Schliesslich verweise ich noch auf das Neutrum gauhgava-, ,, Name
verschiedener Saman “, das iiber den Eigennamen Guiigu- letztlich
wohl auch von der y/ gn abgeleitet werden muss und sichtbar die
Verbindimg mit Gobhila-, dem Verfasser eines ziun Samaveda
gehbrigen Sutra, herstellt Mit diesen Einzelnachweisen ist der
vome ausgesprochenen Bedmgung Geniige geschehen, das Verhaltms
^ Das PW. verweist auf Pancavimsabrahmana xiv 3 und Ind. Studien 3, p. -15.
Die Stelle lautet : ,, yadannam vittvdgardadyadaganguyattadgaungavasyagaungatvam •
P\V. zu gauhgava- leitet dies ebenfalls von Guhgu- ab und bemerkt in ,, Verbesse-
rungen und Nachtragen zu Theil i— v." 5, Sp. 1380 unten: . wenn davon ganhgavh
kommt, so ist wohl agunguyat zu lesen aber Caland liest gleichfalls a-gahguyat
(vgl. Pancavira&a-Brahmana. The Brahmana of twenty-five chapters. Translated,
Calcutta 1931, p. 355 ; auch sonst bietet Caland nichts Weiterfuhrendes). Er sagt
nur p. 356 oben : ,, The meaning of agardat and agahguyat (r. aguriguyat ?) must be
guessed eine Ausserung, die mir angesichts des morphologisch klar zu V gx singen
^horigen agardat unverstandlich ist. tJbrigens ist die Stelle genau Pancavim^'
KTahmana xiv 3, 18, 19.
ARISCHE KULTUEGESCHICHTE
871
zwischen gu und *\/ gubh erschopfend geklart und insonderheit
fiir *'\/ gubh ^ und den von ihr abzuleitenden Eigennamen Gobhila-
der Standort im Sprachgefiige endgiiltig bestimmt, und zvar im
Sinne dessen, was ich „ Wortkundliche Beitrage zur arischen
Kulturgeschicbte und Welt-Anschauung. i. “ p. 107 f. ausgefiihrt babe.
9 . Von bier aus wenden wir nocbeinmal den Blick zuriick auf
Gobhila-, dessen mittelindoarische Vorform goha-, m. oben p. 857 f.
als ,, (in Aufputz, Verkleidung auftretender) Sprecher “ bestimmt
wurde, wobei die samtlicben Bedeutungsscbattierungen des Eomens
goha- sicb passend aus dieser Grundbedeutung erklaren liessen (oben,
p. 857 mitte-859 mitte). Leider sind in den daran angescblossenen
Einzeluntersucbungen keinerlei Instanzen sicbtbar geworden, die
es uns erlauben wiirden, fiir altindoariscb zu erscbliessendes *gohha-,
m. und dessen Ableitung Gobhila-, n. pr. ganz sicbere, alteste Bedeu-
tungsansatze zu ermitteln. Soviel icb sebe, kann man trotzdem fiir
*gobha- selbst etwa bei der oben angegebenen Grundbedeutung
bleiben und in diesem Zusammenhang, wegen rebhd-, auch den
Akzentsitz festlegen, *gobhd-. Dass Gobhila- kein Deminuti\-um
dieses *gobhd- war, also etwa nicbt mit „ Sprecberlein “ — so anmutig
das aucb klange ! — wiedergegeben werden darf, scbeint mir aus
dem, was p. 839 f. iiber die Rajputen-Namen auf -e,f7 und p. 861 f.
iiber die Funktion des b)'pokoristiscben -ila- Formans gesagt worden
ist, hinlanglich klarbegriindet. Aucb litauiscb tirszkalas ,, Scbwatzer
und abd. wortal ,, gespracbig welcbe beide in einem unverkenn-
baren idg. Bedeutungs- und Form-Verband mit Gobhila- steben,
sind nicbt deminutiv. Gobhila-, als Name eines Sutra-Verfassers
aucb sacblicb passend gebildet imd ganz sicber nicbt zufiillig
* Den Grand des Erldschens der *v' 9 »bh sehe ich 1, in der sie umgebenden reichen
Synonymik (\/-\/ gad, gd, gr, rihh, vac, rad, rand, slu, stubh usw.) sowie 2. im Erloschen
der sie stutzenden V git. Der idg. Ansatz ist keinesfalla ‘•y/ ghcnhh, sondern
* V goubh wegen -y' goii (Walde-Pokorny 1, p. 634 f., aber ohne jeden Hinweis
auf die Moglichkeit einer -bb- Erweiterung). Theoretisch moglich ware *-y/ gheubh
nur mit Wirkung des Grassmann'schen Hauchdissimilationsgesetzes, aber dieser
Ansatz wird durch alles oben Ausgefuhrte, wie betont sei, dringend widerraten.
Deswegen irrt auch F. B. J. Kuiper, Zur Geschichte der indoiranischen s- Prasentia
( = Acta Orientalia 12, p. 190-306), p. 268 unten und f., wenn er als Parallelfalle
„ ved. stobhate zu V stu, sobhate zu su- (vgl. sue-, sudh-) “ anfiihrt, aber dann der
einzig moglichen Schlussfolgerang auf -y/ yweinfach ausweieht. Altpersisch y ' g(a)nb
ist also unter gar keinen Umstanden eine Stutze fur die Kuiperschen Konstraktionen
zu einer idg. *\/ gheus . — Was unsere *y/ gubh glottogonisch-morphologisch ist,
geht uns hier nichts an. Es liesse sich denken an -y/ XuUstufe ii der -y^ bhd,
also an einen ungefahren Typus wie etwa lateinisch condere. Doch ist dies nur eine
Vermutung wie ebenso die semasiologische Proportion : y/ subh : *y' gubh = y^ bhd :
latein. /on.
87-2
W. WUST
Samavedist wird im Eigeimamen-Typus etwa an altindoarische
Fomina propria mit dem Ausgang °-vdglsa- oder °-vdcaspati- ange-
schlossen werden diirfen AIs zweckmassigste, sinngemasse Wieder-
gabe erscheint mix das nhd. Wort ,, Barde weil dies die im Namen
Gobhila- zusammenfliessende priesterlich-magische und beroisch-
kriegerische Sphare nicht unwirksam veranschaulicht. Ob die bekannte
Eegelung des Apastamba-Gesetzbuches (i, 3, 10, 19), dass das
Yedastudium dort unterbrochen werden miisse, wo Hundegeschrei.
Eselsgeschrei, das Heulen von Wblfen und Schakalen, das Schreien
der Eule, der Klang von Musikinstrumenten, Weinen und der Ton von
Samans gehort werden, auf den Sinngehalt von Gobhila-, insbesondere
iiber die Grundbedeutimg des Nomens goha-, Bezug babe, muss eine
offene Frage bleiben. Sie kann es getrost bleiben, denn der
giiltigen Ergebnisse sind wahrlich genug. Wir fanden zwei Satze
J. Wackernagels bestatigt, mit denen ich gerne schliesse : ,, Dass
trotz allem, was die modernste A’edaforschung gegen die Verwertung
der Sprachvergleicbung fiir das Wortverstandnis des Veda ein-
zuwenden hat, eben dock sogar sonst undeutbare indische Personen-
1 Vgl. Knauer, 2. Heft p. 52 f. und hiezuerganzend H. Oldenberg, Sacred Books of
the East 30, p. 3'.
2 Ein freundlicher Zufall tritt in dem iranischen Eigennamen eines Hichters,
Gufti, zutage (Justi, Iranisches Xaraenbuch p. 119, Sp. 1, p. 496 f. bemerkens-
werterweise nichts), der eigentlioh ,, Rede, Sprechen “ bedcutet und von der gleichen
Wurzel gebildet ist wie altinfloarisoh Gobhila-. AUe iibrigen idg. Spuren sind frag-
wiirdig. Lcttisch gauhju, gaubt ,, sich ergotzen. jul)eln “ : trotz der sehr passenden
Pedeutung, wozu noth ,, versprechen, .sieh beklagen ‘‘ kommen, kaum verwertbar ;
P. Persson, Beitrage zur Idg. Wortforschung p, 59, Walde-Pokorny 1, p. 567.
Muhlenbach-Endzelin 1, p. 694, Sp. 1 f. s.v. I gaubt. Endzelin, Lettische Grammatik.
p. 114 unten und p. 131 oben ; Lehnwort au.s dem Mittelnicderdeutschen wegen
— Mhd. guft ,, clamor “ u.sw. : dies wohl <; idg. *gheup ; Hiefenbach, \gl.
Worterbuch der gothischen Sprache 2, p. 5.>4 unten (mit keltischen Wortformcn).
.'^ehade, Altdeutsches Worterbuch 2, p. 3.56, Sp. 2 f., p. 3,57, Sp. 1 f., Walde-Pokorny
1. p. .567 : hubsch i.st der .Sirenengleichklang mhd. guften „ prahlen “ : neupersisch
giiftdn ,, sprechen *■. Xichts als Trug ware es auch, bei Gobhila- an den nhd. Familien-
namen Gobel denken zu wollen (etwa mit niederdeutscher Lautgestalt) ; Gobel
vielmehr < ahd riobbilo und zu Gotlebald (v’gl. Solmsen-Fraenkcl a.a.O. p. 177 unten
und f.) ; fur derartige Sirenenklange i.st E. Littmann, ZDMG. 76, p. 273, nach-
zule.sen. — Am ehesten ware noch zu erwagen Zusammenhang mit spatlateiniscli
gufo, m. ,. Eule : A. Ernout. Les elements dialcotau.v du vocabulaire Latin
(= Collection dc linguisticiue publice par la Societe de linguistique de Paris — iii).
p. 131 f. (verbindet altpers. gaubataiy und schliesst auf -bh) ; Thesaurus linguae
Latinae s.v. ; G. Landgraf Archiv fur Lateinische Lexikographie und Grammatik
9, p. 367 f. ; W alde-Hofmann, I-ateinisches etymologisches Wfjrterbuch® p. 62.5 ;
55 . Meyer-Lubke. Romanisches etymologisches 5\'orterbuch^ p. 333, No. 3908 ;
Oorpus gloasariorum Latinorum vol. v. Placidu.s lil>er glos.sarum. Glossari.a reliqua.
EOidit Georgius Goetz, p. 272, Zeile 40/41. Bei Zusammenstellung des letztgenannten
™'r Dr. P.aid-Max Groth, wissemschaftlicher Assistent
.Seminar der Universitat Munchen. dankenswerter-
AEISCHE KULTURGESCHK'HTE
873
namen auf diesem We^e Liclit pmpfau<reii. majf auch der Xaiiie “
Gobhila- gezeigt habeii. Aher ,, iiu allgetneiiieii lolnit es sich nicht.
der Etyinologie von solchen Personetitianien nachzugehen. denen
aus dem sonstigen Wortschatz nichts zur Seite steht " (Indoiranisches.
Berlin 1918, p. 405, Ahsatz 3 und 2).
10 . Aufriss des Hauptergebnis.ses (gemass deni von mir
so genannten Synthetischen Lexikon-Schema " oder dem
Zusammenfassend-darstellenden Donkbild in Worterliucliform “) :
Gobhila-, ni.n.pr., Yerfasser des nacli ihm benannten Gobhila^
grhyasutra. Erforscliungsgeschiclite, Gesamtbeliandliing und Losung
bei \YaIther \Yust. AVortkundliche Beitrage zur arischen Kulturge-
schichte und Welt-Anscliauung. ii ( = Bulletin of the School of Oriental
Studies, University of London, volume viii. parts 2 and 3. p. 835-73).
London, 1936.
Seit dem spiiten Veda + (Stellenbelege z.B. im P\V. s.v.). —
Gobh-ila-. im Klang- und Form-Verband mit Oinila-. m.n.pr., Somila-,
ni.n.pr. und RMiila-, m.n.pr., ist Hypokoristikon zu mittelindoarisch
goha-, m. (in Aufputz oder Verkleidung auftretender) Sprecher "
< altindoarisch *gobhn-. in. (vgl. rebln’i- : Ri bhiln-). Gobhila- mag etwa
so\del iN’ie .. Barde " bedeutet haben. Das Wort lebt fort im
prakritisierten m.n.pr. GohiUa- .sowie im Rajputen-Xamen Gohil.
Eigenstandig-innersprachliche Schiclit des Alt-Indoarischen, wa.s
die Wortform als Ganzes anlangt. Die Wurzel ist = altiranisch
(altpersisch) y ganb, gab .. ilicere ",
Ableitungen : gubhiUga-. adj. .. zu Gobhila in Beziehung
stehend". -rgobhiliha-, ^gaubhiUka-. beide Adj, patronymica. //m(6/n7a-.
n. ,, Gobhila.s Grhvasutra (vgl. Saamila - : Somihi-). Wichtigere
Zusammensetzungen fehlen. Eine Deutung der einheimisch-indischen
Sprachwissenscliaft ist nicht bekannt.
Verfehlt oiler teihveise verfeldt ; .Tame.s Tod. Annals and antiquities
of Rajasthan ... I. ]). 137-; Fitzedward Hall bei Henry M. Elliot-
.John Beames. 41emoirs on the lii.storv. folklore, and distribution
of the races of the Xorth-We.stern Provinces of India 1. p. 91 f. (und
p. 90 2) ; Albrecht Weber. .Vkademi.sche \'orle.sungen itber Indische
Literaturgesehichte' p. 92 unten : M. Bloomfield. The .Johns
Hopkins L'niver.sitv Girculars. Xovember 1882 -Octolu'r 1883. p, 141 f. ;
Jarl Charpentier. IF. 29, ji. 380 f. : Irach .1, S. Taraporewala. Indo-
Iranian studies ... in honour of 8ham.s-Ll-Ullema Dastur Darab
Peshotan Sanjana (London-Leipzig 1925). p. 14.')-8. besonders p. 147
unten und f.
-> \/gu.
VLiT,. YTTI 4\n ” '
Two Yazghulami Texts
By I. I. Zarubix
A MONG the specimens of the Pamir languages included by
Sir George A. Grierson in his Specimen Translations in the
Languages of the iSorth -Western Frontier (later on republished in the
Linguistic Survey of India, vol. x) the YazghulamI is not represented
at all. In his later work specially comprising the Ishkashmf Zebaki,
and Yazghnlnmi (London, 1920) Sir George A. Grierson, having no
texts at his disposal, hail again to limit himself to a mere li.st of about
thirty words in YazghulamI tliat were taken down by Sir Aurel Stein.
R. Gauthiot, too, because of the .short duration of his investigations,
could not spare time for the taking down of texts, although in his
'■ Notes sur le Yazgoulanii " (Journ. Asiatique. 1916) he gave the first
scientific (mainly historic and phonetic) characteri.stic.s of that
language. Thus until a few years ago there remained only one
published example of the .spoken YazghulamI. i.e. the three lines
quoted by C. Salemann in his Maniehivische Stndien in 1908. After
that it was W. Lentz who. in 1933. included a p()em consisting of five
couplets in Yazghuh'uni into his Pamir-Diahkte. Therefore the
publication, even if a belated one. of two YazghulamI texts can still
form a natural supplement to Sir George A. Grierson's summary and
be of use for the comparative characteristics of the Pamir languages.
Both texts were taken down in Augu.st. 1915, as dictated by
Sufi, a middle-aged inhabitant of the large village Matraun (Yazg.
MarOtin), near the confluence of the Yazghulam river (Yazg. Yuzddm)
and the Pandj. The texts were taken down in the presence of several
of the local inhabitants who showed great interest in the proceedings
and made remarks of their own. The texts are not original ones ;
they are stories rather popular in Central Asia, which were originally
taken down (also in 1915) in Shughnl, with the help of an inhabitant
of the village PorshnB' on the Pandj. whose name was Ghulam-All,
in the presence of and in participation with Said-Sha-Fazil, of the
same village.
In order to facilitate the right understanding and comparison of
the text.s the ShughnI original is placed next to the YazghulamI
version.
876
I. I. ZARtJBTX —
The transcription used is the one employed by R. Gauthiot in
his dialectological papers. The only deviation from it is made with
regard to labialized velars. R. Gauthiot being somewhat inconsistent
as far as their transcription is concerned ; this will be seen from the
following examples : " A-a«-. kiin- faire, scear, <f>uoi'd lait.
xwaylrg moulin a eau, xu'ar manger." In order to represent those
sounds which make the YazghulamI sy.stem of consonants differ
from that of the other Pamir language.s more consistent and usual
characters are employed here : A’®, P®, y®, x® ; for instance :
I do, h'^anit you do, k^ana) to do. .c”’an blood. y’-^o9 excrement. Jf’^ant-ai
he read, x'^drd eats, tax'll burnt.
A more detailed analysis of YazghulamI is soon to appear in an
edition planned by the Institute of Language and Mentality of the
U.S.S.R. Academy of Sciences, comprising all the YazghulamI texts
that were taken down in 1915. Therefore it seems more suitable to
communicate here only those of the peculiarities of YazghulamI
morphology, which are altogether necessary for the understanding ot
the texts and which can form a complement to data that are already
known.
Already Gauthiot noticed a difference in the treatment of transitive
and intransitive verb.s. There are in connection with it some
distinctions with regard to the verbal paradigm of the past tense of
transitive and intransitive verbs :
Intrans.
Sg. 1. dz-am /'rapt I reached.
2. Idw-af/'rapt thoureachedst.
3. mf. ai (or u) f'rapt he, she
reached.
Trans.
Sg. 1. iwMw/'rapd/iI (-am) I caused
to reach.
2. tu /‘rapdnt {-at) thou
causedst to reach.
3 m. dai (or wai) f 'rapdnt (-ai)
he caused to reach.
f. dim (or im) /‘rapdnt (ai) she
caused to reach.
PI. 1. mox-an /‘rapt we reached.
2. you reached.
3. di/ (or i/)-an /‘rapt they
reached.
PI. 1. mox /‘rapdnt (-an) we
caused to reach.
2. tamox /‘rapdnt (-a/) you
caused to reach.
3. dif (or l/)/‘ rapdnt (-a/) they
caused to reach.
T\\0 YAZGHULAMi TKXTS
877
The pronominal suffixes which are here put in parentheses are
usually omitted, being employed only when the subject is missing.
In the present tense the transitive and intransitive verbs are
conjugated alike, the particle -da -la -a. characteristic of the present
tense being added ; —
Present
Intrans.
Sg. 1. dz-a f‘reptn I reach.
2. toiv-da f’repdi thou
reachest.
3. ai (or u)-da f'rept he, she
reaches.
PI. 1. mox-ta f'repxm we reach.
2. t»m6x-ta prepit you reach.
3. dij (or yyta f'repdn they
reach.
Tense
Trans.
Sg. 1. dz-a farpanin I cause to
reach.
2. toir-da prpandi thou
causest to reach.
3. ai (or u)-da fxrapdnt he, she
causes to reach.
PI. 1. mox-ta farpanhn we cause
to reach.
2. tamox-ta farpanit you cause
to reach.
3. dif (or ifyta farpandn they
cause to reach.
As can be seen from these tables, there are special forms of pronouns
in the oblique case, but only conserved in the singular, and only in the
oblique case of the singular can there be observed a different form for
the feminine gender. The pronouns of the 3rd pers. are. at the same
time, demonstrative ones different for the proximate persons and
objects [ai this, obi. c. dai. fem. dim, pi. diJ) atid the remote (u that,
obi. c. irai, fem. im, pi. If). The other nouns have not kept this
distinction between the direct and oblique ca.ses and express the case
relations by means of prepositions and postpo.sitions. Only when
being used as an attribute can the noun he employed in a special
“ attributively possessive " form, obtained by means of the affix -i,
as can be seen from the following examples : i, 7 : tfi caUfiri vred
their younger brother ; i, 51 : uai Jcaldftri 8nyd his elder daughter.
At the same time, the putting of the attribute after the noun, so
characteristic of the Persian language, frequently occurs as well ;
therefore in the text the ” owner of the camel ' is ccjually e.xpressed
both by s<Pihi axtor and by axtdri xo'^ih.
878
I. I. ZARfBIX —
These remarks, together with the parallel Shughni text, a trans-
lation into English, and the vocabularies already published by
Sir George A. Grierson and R. Gauthiot. seem to be sufficient for the
purpose of allowing these Yazghulanil texts to be read and used by
specialists.
YAZGHrLAiii Vkrsiox
I
1. vad-an yo na vad-an. ar cui vr4d-an vad. 2. Sou v^red-an na wu
nan, 3. wug na wu nan. 4. Tf-me-n cui wus vad. 5. cigag-da xud paid.
6 . koldarad-a xud na payan. 7. wu mid ffi caldflri vred xe wus na
peid. 8 . wai vradarwai wuszfid. 9. u 3 ’atko: ni wus na vast. 10 . bat-ra
xe wus xikard. 11. xe wu.si'nuirSa-i v'vug, 12. zgxt-at .sod I daxt. 13.
tamox 6 -i keg ke wu axtor .s '-pred yat. 14. yaSag axtor n'yiid, wusi
xVin-at mi y^od axtori yS.x-ra maid. 1-5. ku.xt-ai ke : soJi'bi axtor vat.
16. so-ibi wus d^rf yarawa| Sed, 17. e vred. laft-ai ke ; axtor ni wus
xug. 18. laft-ai ke : e vred. daroy na laf. axtor-da na wus na x^'ard.
19. laft-ai ke : mi ya.x kas. ke mi y’^od-at mi x'^an maldagin. 20.
az-da badin p^’ta qozl. da arz k'^anin. 21 . axtori so5'ib laft ke ; arz
na bad, ni axtor ti wiisi qarz-ai. 22 . du yaSag nia axtor s'wfr mad-at
.sod. 23. wai vradar kuxt ke ma axtor s^wer-at vat. 24. laft-af ke :
e badbaxt. tu yu a.xtor na.s-kore-at vag ? 25. ay-ai kf-rne ? 26. wai
laft ke : ni wus-af zud, 27. mun mi murSa va.st, 28. axtor-am zext.
29. If xe wus zud, mi murSa avjd-af d^ri bozor. 30. mardum-da pIsan
ke : av-ai cig ? 31. If laft ke : ay-ai wii.si murSa. wii.si murSa-da
axtor-itu SaSam. 32. mardiim laft ke : e beaqlad, wiisi murSa-da
axtor-itu ki nh’ast ? 33. if-an na xe kar xarminda mad-at s 6 d-an.
34. xe wiis-af zud. na-f a.xtor vag-at na wus. 35. bat-ra-f masalhat
keg ke : dai badbaxt z-mox ibriit 5«"d, 36. mox-ta dai nan zanam.
37. bat-ra-f dai nan zud. 38. ai .saylra mad. 39. xe nani murSa ma
xur s^’wer keg, 40. yarawarm 3 'at wu jinda. 41 . wu wexiig kaxta-da
zazd. 42. li n'viist, 43. xur .sod dYi ka.xta. 44. .soGbi kaxta yar Sed
xfirag. 45. imi murSa na xur wob Sed. 46. vtike yaSag vat, pai yarawa]
mad. 47. laft-ai ke ; ni nSn-at ziid. 48. ta qozi-da badin, arz-a
k''Anin. 49. laft-ai ke : mu-me Sou Soyd, 50. wug xu.s kan-at zaz am
xe qarz. 51. du yaSag wai kalduri Soyd zgxt-at sod I kfid. 52. wai
vradar laft ke : nas-kore-at vag ? 53. laft-ai ke : xe nani murSa-am
Sed, \m yacag- 8 m idyud. 54. If-Ja xe nan zud, avM-af d^ri bozor.
55 . inarduin z-if pl.st ke : vu niurSa-da .satii k'''anft ? 56. If laft ke ;
asA yacag-itu-da SaSain. 57. mardum if-ra izzo Sgd-at .s 6 d-an I kild.
TWO YAZGHULAMi TEXTS
879
II
1. wu mI0 qozi kitob x’^ant. 2. kitob-and-ai wint ke : bar odame
ke wai kal-u cigag, wai ben-u qatol, nak-dii odam beaql. 3. bat-ra-i
kuxt ke : wai kal-u cigag. wai ben-u qatol. 4. xe zawS-and-ai laft ke :
xe kal-da qatol k'^ana| na bas-sain, xe ben-da mes k'^anin. 5. bar
ciind-ai naqroz-talapt ke, na vVug-ai. 6. s5m-ai xe Siist-ama xe
ben n’Tt’nd. 7. sam-ai xe bfn ayed ta c^rSy. paSaid-ai. 8. wai ben dar-tob
Sed, wai Sust tax"!. 9. xe Sust-ai Sar keg. 10. xe bSn-ai lai k'eg. u
pogiza tax^'t. 11. qozf xarminda mad ke : 12. bar ci ke wint-ni
razg vad, ata ay-ai daray.
Origixal Shughni Texts
I
1. vu8] na vu8J, arai vr6d-en vaSj. 2. Si'iin vr6d-en as yi nan. 3.
ylw as yi nan. 4. wev-and-en arai sig vie. 5. jub'k xund poid. 6. waS
katanakin xund na p6yen. 7. yi me^at yu mis xiind-e na peid. 8.
waS we \Todaren-en we sig zid. 9. yu yat dide : we sIg nist. 10. daS-e
^ik-amard xikud. 11. we murSa-e vud. 12. zd.xt-e xu. sut tar diixt.
13. tamos6-e cud : yi xetur as per6 yat. 14. yu daS wam-e anjfivd-e
XU, sig wexen-at we yad-ri.s mfilt-e warn xetxir yev-ard. 15.
ciixt-e : warn xetiir so'i'b yat. 16. yam sigdbr daraw-e nlwd .sut.
17. Ibd-e : e vrod, tu xetur mu sfg-e xud. 18. yulfid-e: e vro. des ma
liiv : xetiir ta as sig na xirt. 19. yu liid : dam yev cis, wexen-at
ya^risin. 20. wuz podxS xez sam doS kinum. 21. yu xeturdbr lud-e :
e vrod, doS ma sa : mu xetiir tu .sig qarz-ande. 22. yu yeSa warn
xetiir-e sawor siit xu. tflid. 23. we vrodSr-en cQxt dide : wev vrod
xetiir saw6r-at yat. 24. lud-en ; e badbaxt. dam xetiir-at as ka vud ?
25. yid cind ? 26. yu lud-e : mu sig-et zid. 27. wiiz-um we murSa
par-§6d. 28. xetiir-um zoxt. 29. waS-en mis xu sag6n zid xu.
wev murSa-jen vod tar bozor. 30. mardiim-en pexst : yid ciz ?
31. waS-en liid : vain sig murSa. xetiir-te SaSam. 32. inardiim-en
liid : e beaqleygn, .sig murSa ta ivOr xetiir-te na zezd. 33. waS-en
a.s xu kor sat xarmenda xu. toid-en. 34. xu .saggn-en zid. na-yen xetiir
vud, na sig. 35. daS-en tar yak digar niuslat cQd dide : yid badbaxt
ma.s-e abrat Sbd. 36. mas denanzinam. 37. daS-en we nan zid. 38. yu
sayira sut. 39. xu nan murSa markab-te sawOr cu. 40. niwd xu, yat
yi jo-ndir. 41. yi corik serzezd. 42. yu nust. 43. markab sat tar ser.
44. yu corik markab 86d-e (limb. 45. yu murSa as wiim-te we.xt. 46.
yam yeSa yat, daraw niwd sut. 47. lud-e : mu nan-at zid. 48. wuz sam
podxO xez, doS kinum, 49. yu liid-e : miiud-eii Su rezin. 50. yiw xu.s
I. I. ZARIBIX —
880
ke XU, zez xu qarz-andir. 51. yeSs we xedar rezin zoxt xu, tuid.
yat tar xu cid. 52. we vr6d5r-en lud ; dam-at as ka vud ? 53. yu
lud-e ; xu nan murSa-ni Sod, mam-um zoit. 54. waS-en mis xu nan
zid, warn murSa-^en yod tar bozSr. 55. mardum-en pexst ; dam
murSa ca kinSt ? 56. waS-en lud : mam basand y5c-tlr par-SaSam.
57. mardflm-en wev Sod izzO xu, wfiS-en toid.
II
1. yi med qoze kitob xeid. 2. kitob-ande wint-e ; '’ar cind julik kal
ved, daroz bun. amaq yik-u odam. 3. bad qoze cuxt-e : wind julik
kal-at daroz bun. 4. xu z5rS-ande lud dide : xu kal katanak cid na
var-Sim, atS xu bungn kut keniim. 5. bad ar cund nuqrSz telapt.
nuqrOz na vud. 6. bad xu bun?n-e kand xu Su.st-and anjuvd-at 7.
we kand pe cerdw-and yod xu. peSed-e. 8. we bun?n-en dar-tdv sat.
we Sust dud. 9. xu Siist-e cud Sar xu, 10. xu bunSn-e lak cud. wdS-en
poyjd peSed. 11. qoze lap xarmandd sut yik-de jat dide : 12. ’’ar ciz
kitob-and ca wint, yu rost na.xtnid-at yam derfiy.
English Tranls.ation
I
1. (Whether) it has been (or) has not (been), there were three
brothens : 2. two brothers by one mother, 3. one by another mother.
4. They had three calves. 5. The younge.st looks after his (own calf),
6. the elder ones do not look after theirs. 7. Once he, too, did not
look after his (calf). S. Those brothers of his slew his calf. 9. He came
and his calf was not there. 10. Then he .searched that place ; 11. found
its corpse. 12. He took it and went to the plain ; 13. he looked : in
front, a camel approached. 14. He took it and with the calf's blood
and chyme smeared the caiuel's mouth. 15. He looked : the owner
of the camel approached. 10. The owner of the ralf .started weeping,
17. said ; brother, thy camel ate mv calf. 18. The other one said :
brother, do not say so ; a camel does not eat a calf. 19. He said :
look at it.s mouth, on it (there i.s) blood and chyme. 20. I .shall go to
the king and complain. 21. The camel'.s master .said : brother, do not
go complaining : let my camel be a compeu.sation for thy calf. 22.
That fellow mounted the camel and rode away. 23. His brothers
looked ; their brother came riding a camel. 24. They said ; thou
wetch. where hast thou brought this <amel from ? 25. ATho.se is it ?
-6. He said ; you slew my calf ; 27. I sold its corpse 28. and got the
"anitl. 2 , 3 . They. too. .slew their calves and brought their corpses to
TWO YAZGHl'LAMl TEXTS
881
the market. 30. People asked them ; what is that ? 31. They said ;
that is a calf's corp.se. we (would) <;ive it for a camel. 32. People
said ; fools, no one will take a calf's corpse and ^ive a camel (for it).
33. They grew ashamed of their behav'iour and went away. 34. Their
calves they slew and got neither camel nor calf. 35. They took counsel
with each other : that wretch swindled us. 36. Let us slay his mother.
37. Then they slew his mother. 38. He became an orphan. 39. He put
his mother's corpse astride on a donkey 40. and. weeping, came to
a certain place, 41. and there a man gathers thrashed grain. 42.
He sat down. 43. The donkey walked on the grain. 44. That man
threw a stone at the donkey. 45. Tlie corpse fell from it. 46. The fellow
approached and started weeping. 47. said : thou hast slain my mother.
48. I shall go to the king and complain. 49. The otlier one said : I
have two daughters. 50. choo.se one and take her as a compensation.
51. The fellow took his elder daughter, went, and arrived home. 52.
His brothers said ; where ha.st thou got her from ? 53. He said :
I gave my mother's corpse and got her. 54. They, too, slew their
mother and took her corp.se to the market. 55. People asked ; what
are you doing with that corpse 1 56. They said : we sell it for a
beautiful girl. 57. People abused them and they went away.
II
1. One day a judge read a book. 2. In the book he saw : whoever
has a small head and a long beard, (tliat man) is a fool. 3. The judge
then looked : he had a small head and a long beani. 4. He said in his
heart : I can't make mv head large, but I will make my beard short.
5. No matter how long he then searched for .sci.ssors, he could not find
any (scissors). 6. Then he took half of the beard in his hand 7. and
brought the other half near a lamp and put it into the fire. 8. His
beard burst into flames and his hand was burnt. 9. He withdrew his
hand, 10. but left the beard. It burnt altogether. 11. The judge grew
very ashamed, because 12. all that he saw in the book turned out to
be true and that (what he had done) was wrong.
Kegan Paul, Trench, Trubner & Co., Ltd.
ORIENTAL AND AFRICAN BOOKSELLERS AND PUBLISHERS
CLASSIFIED STOCK OF NEW AND SECOND-
HAND BOOKS ON THE LANGUAGES AND
LITERATURES. ARTS, HISTORY, GEOGRAPHY,
AND ANTHROPOLOGY OF ASIA AND AFRICA
INSPECTION INVITED
FOR THE MOST COMPLETE RECORD OF NEW LITERATURE
on the above subjects
KEGAN PAULAS LIST OF NEW BOOKS
Catalogue XXVIII. Africa and African Languages,
over 2,000 items, now ready. Sent post free on application
Libraries and collections of books and journals purchased
38 GREAT RUSSELL STREET, LONDON, W.C. I
OPPOSITE THE BRITISH MUSEUM
SUBSCRIBE TO \
ANTHROPOS
International Review of Ethnology and Linguistics.
Founded by the Rev. Prof. W. SCHMIDT, S.V.D. '
Edited, with the co-operation of nurtierous missionaries and i
savants, by the Rev. Dr. GEORGE HOLTKER. !
The greatest scientific periodical dealing with every aspect of Anthropology.
Devoted to studies in Ethnology, Comparative Jurisprudence and Religion, ■
Linguistics, Physical Anthropology and Prehistory.
List of international contributors includes, on the one hand, missionaries and
laymen who are actually working in the field, and, on the other, the specialists '
in the great Universities of the world. |
Published since 1906 in annual volumes of 1,000 pages quarto, illustrated,
with plates and maps.
“ Anthropos ” has published the following articles in :
VoL. XXX (1935), 1-2.
\5ilhel.m Kopper.s : Die ludogermanenfrage im Lichte der historischen
VOlkerkunde .......... 1
Profesob Do.v Fr.wci.sco .M.\ktinez L.vndero : Los Taoajkas 6 Sumos
del Patuca y Wampil (Mapa) ....... 36
Gb:neralarzt Dr. Leo.n B.vlner und Dn. Viktor Leuzki.ter : Zur
Riologie und .tnthropologie der Kenja in Xordost-Borneo (ill., Karten) 51
Rev. Perl Edmond C’esard : Le .Muhaya (L’Afrique Orientale) (ill., Carte) "5
Dr. C. Streit : I’nbewegliche Korperzier in vorgeschichtlicher Zeit . 107
Rev. S. Gn'.an'.a Prakas.vb, O.M.L : Root-words of the Dravidian group
of languages .......... 135
P, Fr.axz Muller, S.t'.D. : Beitriige zur Ethnograjdiie der Guaraui-
Indianer im ostlichcn Waldgebiet voii Paraguay (ill.) . . .151
Rev. Pere Jo.seph Kler : Quelques notes sur les coutiimes matriinonialcs
dcs Alongols Ortos (Urdus) Sud (ill.. Carte) ..... 16j
Dr. Herbert Bai.dus : Sprachproben des Kaingaiig von Pahna.s . 101
WiLiiELii Schmidt : Fritz Graehner t (Portrilt) . . . . . 203
.\nalecta et Addjtarnenta (215), Miscellanea (229), Bibliographic (251),
.\vis (325), Zeitschriftensehau — Revue des Revues (331).
“ Anthropos ” should be in every scientific library !
Annual Subscription, price Mk. 42. —
Of the earlier v'olumes (many of which are now available only in limited
numbers) the following can still be obtained: — Vol. I (1906), VI (1911), VII
(1912), VIII (1913), IX (1914), X-XI (1915-16), XII-XIII (1917-18), XIV-XV
(1919-20), X\T-XV1I (1921-2), XVIII-XIX (1923-4), XX (1925), XXI
(1926), XXII (1927), XXIII (1928), XXIV (1929), XXV (1930), XXVI (1931),
XXVH (1932), XXVIII (193.3), XXIX (1934), XXX (1935).
Price; Vcl. I, C,0 Mk.. Vols.VI-IX, 80J/ft., Xl.XI-.X.XIII, 52Mk., XXIV-
XXX, 42 Mk.
Prospectuses gratis on reqitest.
VERUG DER INTERN. ZEITSCHRIFT „ANTHR0P0S“,
MODLING b WIEN, ST. GABRIEL, AUSTRIA.
BULLETIN
OF THE
SCHOOL OF OEIENTAL STUDIES
VoL. VIII. Part 4.
Ttaugara
By H. W. Bailey
O TEIN MS. Ch. 00269 is a report by hada (‘ envoys '). who call
^ themselves nam blsa {' humble servants '). on behalf of seven
rispura {" princes '). It consists of 120 lines, but of that a considerable
part cannot yet be translated. The report is concerned with the cities
of Sacii and Kaiiimicu. Danger from robbers is reported and an
investment of the city of Kanimicu, so that food and cattle are lacking.
Then comes the following passage, which is quoted here for the
geographical names (lines 75-80) :
75 cu jsain karnmicu badiimna tta vana drrai pacada stare
76 cu ttudlsa u tturkibayarkava u hattibara u
ici liiijuva ' tta ylpikirnnittahi ; u bedi 77 dartiki
nasta stare
ci buri hvaihu'.ra ^ tta bisi karastahi: nasta
u karastaha
78 cu dumva u cahi;spata u sfdya ® tta jsain kTtha
khu tta viiia drrai pacada bisi hani 79 tsa ni samlnide
kammicu ha hervi * hadi ni ttranidi hame
salai aphaje ® u salai jsani va 80 khiiysi nisti
^ u it'i Inijuva is written under the line.
- hvaihu-.ra are probably Vighurs, of. Tib. //or, JRAS 1931, 832, and Chinese
gj ^ yndi-yuit.
^ sulya seems elsewhere, in a document from the K.hotan region, to mean the people
of Kasyar (Tib. su-Ug).
^ hfiriJ, often in this document, ‘ any
aphaje ‘ investment (?) Cf. phaj- in hnmpkaj- ‘ to envelope ’ ; rKi^phaj- occurs
- YiW
in'
884
H. W. BAILEY
Translation :
As to the baduna ^ in Katnmicu, three sections now exist. As to
ttudlsa and tturMbayarlcava and the hattibara and id {and) imjuva,
they are now settled in yipilcimnittahi: and bedidaruM. As many
as are hvaihicra, all those are settled in harastahi: [and harastaha].-
As to dumva and cahiispata and sulya, they are now in the city, so
that now all the three sections disagree together. No envoy has entered
Kammicfi. On the one side is investment {?), on the other side there
is no food.
It is necessary to confront this with a passage of the Sacu document,
published by F. W. Thomas and Sten Konow, Two Medieval Documents
from Ttin-huang, 27-31, which contains a list of names, probably at
the same time names of places and of peoples. The spacings are
significant.
inijuva yahiida kari adapahutti baku
basikatti karablri karibari tti tfillsi stare
• ■ imju si kari ttaugara ayabiri
caraihi; yabutti Imri anahi:dipabhutti
karattaha ® pata tti ttari ttrusahuta. sadimlya *
ttrrukibayarkata ® cunuda.
The first list is certainly concerned with Karnmicu near Sacu
(— Tun-huang), which is doubtless Mm tsdu (kan-tsou,
kan-chou) in Kansu. The name occurs also in Stein MS. (Brit. Mus.)
5212. 4 hamacu klthdsta ' to the city of Kamacu and in the Sacu
Document 16 kamm-acu kathi u laicu kamtha u sahvd kamtha.^ In
’ hrVluinna ocriirs in two other pa.s^age.s of this document : 6-1 tti mini bisi
haitenitini iirrt ttyntn pnhitfiam "ulatnfU .v/ . . . ' tiius all the haduna attributed {ida-
partir. to ar- in ham-nr-. hamula ' join ' and namavandn ‘ famous ’) the fault to those
patimvi. saying that .... 69 rr ktirni hndumna ttu fieri bausta ' and when the baduna
understood that matter '.
- A[)parently <litt<igraphv.
^ karifttiiha perhaps corrc.spond.s to ka ra-itaha of the iirst list, pata could represent
Ohl Iran, pnti- ' lord of. Khotan .Saka njx'tta, upa ' general ’ <c *spddapati, corre-
spoufiiug to Tibetan sdr-dpon. So read (in place of pata. yd) spdtn, spa in the Sacu
doi umcnt 47. .39.
^ xidimiya is struck out. One might think ofSolmi.
tfrrdki-. ftiirki- of this name is perhap.s ‘Turk’, Tib. dru-gu (F. W. Thomas.
1931, 816 if.). In thi.s .same document Ch. 00269, line 48, we find tturfei 'uhai
hiri ■ belonging to the chief trj of the Turks '.
^ latrii is almo.st certain. In Ch. 0048, 5, karnmicu and meu are also mentioned in
association. In the same context occurs svahva, where one will no doubt recognize
sdhvH.
TTAUGARA
885
the two lists tudlsa corresponds to tuTisi} tturkihmjarkdva to ttrruki-
bayarkata (so to read), hnjum to ijura. We shall probably infer that
the other places or peoples are in the same neighbourhood. This is
important for ttaugara, which is thus assured about a.d. 800 in the
region of Kanchou.^
This notice of ttaugara provides the indigenous confirmation of the
oft-quoted passage of Ptolemy, Geogr. vi, 16 (see F. W. Thomas,
JR AS 1931, 834-5, A. Herrmann in Sven Hedin, Southern Tibet,
viii (1922) 212) who names from the itinerary of Maes Titianus the
people dayovpoL, and the mountain Oayovpov opos south of the
city of doydpa. It is clear that at that time doydpa was an
important city on the silk route, and it has therefore been identified
with Kanchou. Since ttaugara seems not to be kanimicii, it mav
represent a small adjacent town or a people in which the name had
survived during the six intervening centuries.
In or near this same region lived according to Tibetan documents
the tJml-kar, phod-kar (F. W. Thomas, JRAS 1931, 834-5, information
which has unfortunately been overlooked by recent writers on the
Tochari). In the same region from the second century b.c. were
remnants of a people called :A: R by the Chinese (most recently
discussed by Pelliot, Tokharien et Koutcheen, JA 1934, i, 37, note 1, —
cited below as Tokh.). The ::A: R; bad in the second century b.c.
passed to the west, and there also a similar name is found : roxapoi.
It is possible, as has often been contended (see the bibliography in
Schwentner, Tocharisch (1935) 18), that the two names, the native
name ^ toyara and the Chinese R R refer to the one people. One
might indeed have expected to find so important a name as toyara,
which persisted for centuries, transcribed in Chinese of the second-
first centuries b.c.
[If the toyara are the R, the conjecture, which has
however no importance for the following discussion, may perhaps
be allowed, after so many earlier conjectures, that the Chinese name
^ For s and i, d and /, cf. sidathasi, sllatkaAa ' Sdtas, Chilas ' in Ch. 1, 0021a, b 15-16,
edited in Acta Orientnlin (in the press). They are probably a section of the ^
t'iet-hk *t6lis.
- Clauson’a very doubtful discussion of this passage setting ttaugara in Tokharistan
need not be considered here {JRAS 1931, 3il9).
® The occurrence of the one name — of doyapa ttaugara thod-knr in the east and of
roxapoL in the west — used of the one people, suffices to prove that this people had
brought the name with them, since the name is found in places too remote for it
to be possible to suppose the name to be a foreign designation. It is therefore evidentlv
their native name.
886
H. \V. BAILKY —
^ K ill its first two syllables is an attempt to approximate to
the foreign name toyara — it is even possible that one ought to pre-
suppose an earlier form *0odyara — at a time, long before Hiuan Tsang's
efforts in the cause of syllabic transcription, when a different,
disyllabic, system seems to have been preferred. It would
of course also be possible to suppose that a familiar name had been
partially adapted to designate newlv-encountered foreigners. In
A.D. 600 -j<z. rras d'di or tdi and ^ was ngi’^ot, earlier d'dd (-d being
inferred from the later -i ; -6 has also been conjectured), and ngitrat
(Karlgren in a letter of date 23.1.1936, and cf. Konow, Corpus
I nscriptionum Itidicarum II, i, 1.x, and ‘ Xotes on the Indo-Scythian
Chronology Journ. Ind. Hist, xii, 6 ; Pelliot, Tokh. 25) : d‘dd-nyiivdt
could perhaps render *toyar- (or *6o{d)yar-). On this theory ^ or
^ would be explained either as a foreign word of urrknown meaning,
possibly a title, since we may think of ^ J sok-ji^ang. Skt. mka-
murunda, if J is here correct, or of P[l} kudt-tiot-sie beside
CiU} kiiat-tiot for the name of Khottal (Chavannes, Documents snr
les Tou-kiue occidentaur, 276) ; or ^ would be a Chinese word in the
sense of ‘ clan the alternative pronunciation of ft {tsie beside zie)
in this name being then perhaps due to dialectal pronunciation in
Chinese and hence replaceable by ^ tsie (which does itself occur in
the sense of ' branch of a family ’). Such a use of ‘ clan ' could be
paralleled in other Central Asian documents. Tibetan or Saka, and
seems to be known in early Chinese. On the same theory the use of
^ ft without which is found in the earliest reference in the
account of the Shi-ki, c. 100 b.c.. where however we find also
beside /]■> ]^ ft. would be due to the two reasons that
the Chinese tended to abbreviate foreign words (and not only proper
names, as is attested by ft ft /g pji (b'ji)- kdju-nji beside nji,
Skt. bhikmni ; ^ ID ddji-ia beside Iji-ia, Skt. arya ; ^ ^
g‘ia-ldm Skt. sanghdranta), and that they could have taken
(■ great ’) as their own word for ‘ great ’. This second reason would
also explain the purely Chine.se invention of the term /J' T1 R
little ft ’ where /J-. siau ' little ' stands in contrast to ’ great .
The name occurs also, abbreviated to before personal names
(Pelliot. Tokb. 40 note 1). ;dz is also used in the name ^ ^ t'di-'C'-'nn
beside ^ alone (JAOS 37, 148), and in t'di-dzhk, 3Iid. Pers.
tdzly ■ Arab ’.
A curious piece of evidence deserves to be cited here. According
to the 31emoirs on the We.stern Lands g on his return journey
TTAUGARA
887
Hiuaii Tsang passed the ruins of Endere (Saca of the Kharosthi
documents, identified by Btein. Ancient Khotan. i. 429. cf. Herrmann
in Sniitheni Tibet, viii, 211), between Xiya and Krorayina. Here,
apparently through the hearing of some tales, he thought to recognize
the ruins of the old (with variants ; read tuo-)^iidlu i) country.
Though the identification may be incorrect, it is evident that Hiuan
Tsang was here using his ordinary transcription of the name of the
western to^an to refer to the to-yara of the east. It may be that he was
purposely replacing the old (now inadec^uate) transcription ^ (^)
by his more exact trisyllabic transcription tuox>i(tld.
It is also interesting to recall that two wooden talilets from Niva
mention a man of the .f| ® ' the n<ii>^'t)l-tsie country see most
recently Pelliot, Tokh. 40. note 1.]
T6y.\RIST.\N, Toxakistax
1. The Same
The detailed discussions of the land of Tokharistiin have now made
it clear that for the period of the fourth to eighth centuries A.u. only
one region (apart from the reminiscence of the eastern toyara near
Kanchou .surviving, as we have seen above, in Greek. Khotan Saka,
Tibetan and probably Chinese) was recognized to have the name
toyara. This was the country between Sogdiana (Su^d) at the Iron
Gates [dar 1 dhanln) and Bamiyan (.sec Mar(juart. Erdnkihr. 199 ff.,
Pelliot. Tokh. 33 ff.). The capital was Balkh {Bag-la in Tibetan)
and the city of Tarmita was comprised in it. It was therefore the old
Bactria. The evidence is furnished by Arabic, Armenian. Tibetan and
Ghincse sources. It is now not disputed.
There remain however certain points to be cleared up - in
connection with the name itself. The Chinese transcriptions are given
by Pelliot. Tokh. 34 ff. and 48 note 1 (cf. Herrmann, Southern Tibet,
viii. p. 450) as follows :
(1) Travellers and Histories
ft taoxudld li Vf ^ t'uox>(dld
M ft ^ tuoxudJd li >X 0 t-uoxudld
M l/C H tnoxadld Rt H fiioxudid
^ Tuis-ho ed. iiOST. p. 94.>. col. 3.
~ It is nccesBarv to treat the problem in somewhat elementary fashion, since
Indiani^ts. who have largely interested themselves in this matter, seem rarely to have
understood the method of writing in consonantal scripts.
888
H. \V. BAILEY —
(2) Eeligious books translated from Sanskrit
0: Wj k‘i '^0 lak
H k‘i”’o Id
I'J' H.
The name is recorded apart from the Chinese in scripts of two
kinds ; (1) using vowels, (2) without vowels. Only the vocalized scripts
can be used to recover the full word.
(a) (1) Greek roxapoi raxopoi, in the east dayovpoi Ooyapa.
(2) Latin (derived from Greek) tochari with MS. variants including
thogarii.
The word occurs in verse scanned Toyapoi, but unfortunately
the verse is late and cannot be pressed to decide the quantity of -a-
(see Marquart, Eransahr, p. 207, note 4). The second form rayopot,
if correct, confirms the scansion with short second syllable.
(3) Armenian t‘uxari-k‘, toxaraslan.
(4) Sanskrit tukhdra, tusdra (in some districts s was pronounced
kh). tiihkJidra, tu.fkdra.
(5) Tibetan in the east thod-kar, phod-kar ; in the west tho-gar,
tho-dkar, tho-kar.
(6) Khotan Saka ttaugam.
(b) (without vowels) (1) Turkish twxfy (ttvyry, twqry) in Manichean
and Buddhist texts.
(2) Mid. Pers. tuyryst’ln].
(3) Arabic-Pers. luxdristdn, ' tuxdrd, also
j'C- (axdristdn and Jt-- taxairisldn (u may represent
u or o), see Marquart, ErdnsaJir, 228 f.
(4) Sjwiac thirrstn.
The d of Sanskrit has been supposed to be due to analogy with
tusdra ' cold ’. The Arabic alif (-a-) and ai (implying w) favour a long
vowel, but are not decisive, since quality may have guided them in
their spelling. Of the vocalized .scripts only Saka is unambiguous
with but most probably Greek also has d. Tibetan does not
normally di.stinguish quantity, and Armenian is unable to do .so.
It is of course possible that the loyara language was indifferent to
quantity of vowels, but d is indicated by the city name (.Sogd.)
8rw")i (*druwdn or *hruivdn, Gr. 0poava)d
^ We have however kr-tcnn beside kwnynk in Sogdian for kroraxjina. I’elliot >
traii.scription *tlarwu>i, Tokh. 31, is naturally not acceptable.
TTAUGARA
889
The medial consonant is either y or (the labial n of the Chinese
forms with yMO is not elsewhere attested : on SjTiac thwrM)). see
below). Saka -g-, pronounced -y-. and Mid. Pers. y in a script which
has distinct signs for g y X' possibly Latin thogarii and the eastern
forms in Greek show y, elsewhere in Greek, Arabic and Armenian
occurs X (to which Sanskrit Ih, hkh correspond). Tibetan {g. k, dk)
is ambiguous. Since yua was used to express the ya of
farynna (Chavannes and Pelliot, Trade ManicJieen retrouve eit Chine
(1913). 208 note), the Chinese also must be considered ambiguous here.
There remains the Turkish tuyry (tu'xrg), which must be a little
more fully treated. It is necessary to protest against the mechanical
reading toyr'i, which since F. W. K. Muller > has had so much currency.
The matter stands as follows. In the Sogdian script which was adopted
and adapted by the Uighur Turks (a description is giyen by A on le Cocp
' Kurze Einfiihrung in die uigurischen Schriftkunde,’ Mitteil. d.
Seminars fur orientalischen Spracken, Berlin, 1919. 93-109). yowels
are not clearly indicated. In Sogdian itself a, ?. u are often left
unmarked, but they may be expressed by the use of * (alif). y. and iv.
Since y. and m may also indicate <1. 7 e. and d u rcspectiyely, only
etymology can decide when • is a or <7. y is i or 7 or c. ic is u or n or o.
We also find ••. 'y. and nv in use. In Sogdian script only consonants
exist. The Turks took and imperfectly adapted this consonantal
alphabet to their yocalic system. Beside the system which left a, i,
II unmarked (there are many Turkish words in which at least one
syllable is written without t!ie yowel sign, as in pyllymz. biltimiz ' we
knew '), it became the custom to use •’ initially for a, • initially for a
(but also, more rarely, a), medially for a or «. y for i and 7 (e nut being
distinguished), ir for u or o. my in the first syllable (but sometimes
only «') for ii or d. So we find in Turkish words tngry tilnyri. ynr yimil,
kntw kantii, yrly yarliy. In foreign words this yowelless system is
ecjually common, particularly for the reason that in many words the
Sogdian spellings were receiyed with the script itself. So we hace
smivne (Sogd. snnvnc) *samananc ' female disciple . ps'k (vSogd.
■ps'k, Mid. Pers. pmsy, Ay. piisu) *pusak ' crown ', srhk. p-b-k. Skt.
' Transcriptions of F. W. K. Muller must lie useil for linguistic purposes with
some caution. He was evidently satisfied to get a set of graphic correspondences
even if he did violence to the phonetic system of the languages. In Iranian his first
attempts to render Persian. Parthian and Sogdian were perhaps e.xcusable at the
time, but they did not give Iranian forms. For Turkish — a .simpler phonetic system —
his method had less evil etfects. But the same mechanical results gave for example
hahy in place of baliq, if the two dots distinguishing q from y were absent.
890
H. W. BAILEY —
sravaka, pr’ty’picd, prtyhjnd Skt. pratyekabuddha, pwtystb, pwtysbt
Skt. bodkisattm, tby’c, t’by’C taiyac ‘ China bx’r, brx'r (Sogd. ^rx'r)
Skt. vihdra, brxm'ddy Skt. hrahmadatta.
Applied to tivyry (tu'xry tu'qry), this means that a great number of
readings is possible. To decide between these readings external
evidence is necessary. This evidence is afforded by the spelling in
vocalized scripts. It proves that three syllables must be read.^ Between
u and 0 in the first syllable it is not possible to decide positively. The
medial consonant can be read y y (or possibly q). The eastern forms
have y in Greek and Saka. and Mid. Pers. tSyaristdn (so to vocalize)
makes y possible for Turkish also. Hence either toyarl or toxar'i (hardly
toqar'i) may be read. A reading toxi'i is purely mechanical and
useless to decide the indigenous pronunciation of the name.
The Syriac thicrstn also needs a note ^ of explanation. It is probable
that the position of the ic is due to a tendency similar to that observable
in Sogdian particularly in the later texts to traject the ic. In Sogdian
Sm’xC- Syic^- is Suxt or Suyd, Mid. Pers. dxxt ‘ daughter ‘ ; rx>t'sn is
roxsx, Mid. Pers. rdsn ' light ' ; suySykt suySlk-i ‘ Sogdians ' beside
syirbyk adj. 'Sogdian'. syicSyxrk siiyhiydnak ‘Sogdian’, and so in
other words. We shall probably read Syriac thwrstn as toxarastdn or
tdx(ira6td)t, or with ii in place of 5.
2. Script
Hiuan Tsang in the i\Iemoirs of the Western Lands M fS
showed himself particularly intere.stcd in the writing and literature of
the countries he visited. He was it seems perfectly familiar with the
Indian Brahml writing, and in Agni,® Kuci. Khasa (Kasyar), and
Khotana he remarks that they used -the Indian writing with some
modifications. His accuracy in this has been attested by the discoveries
of MSS. in Central Asia. But he came upon other scripts which
evidently seemed to him to need further description. Happily he
thought to describe the scripts of Sogdiana and Tokharistan in detail.
^ Tlif -}j of the Turki."!! form !ia» proliahly imt the same origin as tlie -i- in
Armenian It seems to be due in both cases to some Iranian form. In
Turkish It may be tlie -i of the nom. sing. Sogdian, as in yijmkij Sogd. ynnjyy, ytnyyy
BSOS. viii, oSS, i)ut in Armenian this would not be possible : it would there be rather
an adjei tivai -I < -ik\ The Turki.sh form is the same in both E^uddhist and Maniehean
text>. Markwart, F€--tgfthe Szihuyfi (1927) 67, read toxgry.
' IVlliot's interpretation of the Syriac form, T<>kh. 4S, note 1, ad calc., is
unar cept able.
^ So the Wei annals, translated by S. Levi, Le '' Tukharit'n," j). 11, state ; I'ernturp
f.st corniiie celle des Brakmanes.
TTArCARA
891
The close correspondence of the two descriptions can be seen when
they are set side by side. We are fortunate in having recent
translations by Pelhot. Tokh. 48 ff.
Sogdiana
^ M ®
n + S',
M +0 ^0
^ Sit ®
M ^ # Eo
S IS
Taisho edition 2087, p. 871, col.
1. 11. 12-13, Beal, transl. p. 26.
Translation by Pelliot.
Les lettres de I'ecriture sont
pen noinbreuses, constituees par
vingt et c]uelques elements
priniitifs. qui se combinent et
s’engendrent et produisent un
large developpenient (de vocabu-
laire). [Ces gens] ont quelques
oeuvres ecrites dont ils lisent le
texte verticalement.i
Tokharistan
^ m ~
fi ffi
m ±m
^ ^ Wo
g & 1^
IS ?lr
^ Mo
Taisho edition 2087, p. 872, col.
1, 11. 17-18, Beal, transl. p. 38.
Les lettres de I'ecriture sont
an nonibre de 25, qui se com-
binent et s'engendrent, et par
leur emploi s'etendent a toutes
choses. L'ecriture se lit horizon-
talement. en allant de gauche a
droite. Les oeuvres litteraires ont
progressiveinent cru en nombre.
et depassent en ainpleur celles du
sou-U (sogdien).
The statements are admirablv clear. Hiuan Tsang is in both cases
describing an alphabetic system of few letters. It is evident too that
he did not recognize the Indian Brahml script in either of these. The
Sogdian is known to us bevond dispute. Hiuan Tsang's description
is accurate. Happily the coins attributed to the Hephthalites (Junker.
‘Die hephthalitischcn Miinzinschrifton.' .S'BHIT 1930) and the
Kushano-Sasanian coins (Herzfeld. Memoirs of the Indian Archivologkal
Survey, 1930. No. 38). beside their Brahml and Aramaic legends, show
us this script of Tokharistiin : it is the Greek script of Bactria. The
Greek alphabet had twenty-four letters, but on the coins a new letter
b is known representing i. Hiuan Tsang seems here; also to be c.xact,
So evident indeed is this conclusion that it is regrettable that Pelliot.
^ In tho dcscrijition of Sogdiana, the T ang Annals (cap. '2-1. T’ p. i, col. S) state ;
^ *• which Chavannos, J)i>ciiment'< sur Tou-kmc. occiiltntiinx. 134,
rentierecl: (Ce-s <jens) sont habitue'^ d ecrire en Ugnes konzontak^.
892
H. W. BAILEY —
ToMi. 53, should have made an attempt to find the Brahmi writing ^
in this description, without reference to the existence of the Greek
writing on the coins.
There is however one very important additional fact to notice,
although it has not so far been mentioned in connection with the
Tocharian problem. In the Berlin Academy are preserved (at least
three) fragments in the same script as that of the Hephthalite coins,
that is, in Greek script. These fragments,^ which Junker is at present
studying (loc. cit., p. 3), were brought from Central Asia. The writing
of Tokharistan was therefore known in the Turfan region.
It is therefore impossible to escape the conclusion that the Turkish
toyari {toxar'i) applies to this, the only attested, literature of
Tokharistan, and written in the Greek script.
3. Langvui^e
Hiuan Tsang (the passages are conveniently brought together by
Pelliot. Tolh. 49 f. ; used earlier by Stael-Holstein, Izv. Akad. Nauk
1909, 479 ff.) states that the language of Bamiyan was a little different
from that of Tokharistan. In Siynan, although the writing was like
that of Tokharistan, there were differences in the language. In KapisI
the language and doctrinal rules were very different and in Syamaka
{siang >njie) also the language was different. This would suit the theory
* The Brahmi .script (Ii.stiiioruishe.s 48 sounds for classical Sanskrit, and in
Dialec t A there are 10 (i( hi was originalh- distinct from dha there are 11) additional
sicriis. Documents in Dialect A contain .Sanskrit words, and in foreign names also such
sounds as h and k'h are represented. Still other Brahmi signs are used to write Barciuj
(Maralbashi) Saka and Turkish.
- The first notice of these fragments was given by \'on le C'oq, ‘ Kokturkisches
alls Turfan,’ SBA]V 1900, 1049 ' . . . mehrere grossere Fragmente einer Bur.hroHe in
eiiier hi? heiite noch unhekannten semitischen kursiischrift. . . .' F. W. K. Miiller added
a postscript, p. 1061 : ‘ Die „ hi? heute nneh iinbekannie ?emiihelie tursii'schrifl " isi,
u'ie ick inzivisr.hen fentstellen konnie, die Behnff der diephthaliten (albLA, richtig
icohl ikaL*)’" "der u-etuxrn hluniii it
The Hcfihthalites had occupied Tokharistan about A.ii. 468. Presumably they
adopteil the Tocharian vriting, since according to ,?uong Jiudii, they had no
w riting of their own (Chavaniies. BEFEO. 1903, 404 ; dan? ce pnyn, oh ne connnit pa?
d'ei ntiin). The script of thc'se Central Asian fragments was in any case not con-
tined to the Hephthalites. Hiuan Tsang records it also in Siynan and ^ ^
slang iiijie, Syamaka, which {cf. Herrmann in Southern Tibet, viii. 447) was Mastfij
and Citral. It was therefore premature to call these fragments Hephthalite.
* One must of course in this piroblem keep in mind the possibility of a phonetic
change of ej, to e cchich would suit the XPers. form haitdl.
TTAUGARA
893
that the language of Tokharistan was Iranian.^ It is therefore to be
noted that the Hephthalite coins use an Iranian titulary BA TO
■ god OZOPOBAAI ‘ hazarapati (leader of a thousand, but long
established as an important rank in Persia) and employ the Iranian
gen. plur. -dm (Junker, loc. cit.).^ The significant presence of y or
y in the name loyara-toyara would be compatible with an Iranian
origin. In the east similarly the city name Qpoava Sogd. Invi}
hnv’m shows a fricative (the 6 of Opoava and the Sogd. S together
exclude t ox V in Chinese the name is ^ tudn y'^dng and tg M.
d'lon y'^dng with t and d').
The possibility however that the Tochari in their wanderings had
changed their language must not be overlooked.^ The R \'ho
remained in the east (spoken of as the /Jn U R) spoke a language
like that of the Khang near whom they lived (see Pelliot, ToWi. 37,
note 1). Strabo (first century B.c.) writing a hundred years after the
Tocharian invasion, knew that the Bactrians and Sogdians spoke
similar languages (ed. Jleineke xv, 2, 8) : € neKTuver ai he rowopa
rij; Apcavrj^ pepovs to’o? xai UepuCiv Kai Mrjhwv xai eri^
TcSe TTpdj dpKTOv BaKTptoyv xai Uoyhiavwv ' eiai yap ttio? xai
opoyXojTTOL Trapd pixpov. Bactra was the Tocharian capital. It
remained one of the capitals of the Hephthalites,^ although their
winter residence was, it seems, elsewhere, in old Baday^an
(Herrmann, Asia Major, ii, 576).
4. Literature
Direct evidence for the existence of this lost loyarn literature is
contained in Turkish colophons from Central Asia. In colophons of the
1 Muslim authors, Ibn al-Muqaffa' an<l Muq.addasl. <iuotMl by Marqunrt. hinK.,nhr
SS-9, indicate that a dialect of/hrM' " IVrsian ' ^as imcl in Balkh from the eic i h
century; probably, as Marquart thought, the language ot JSa&anian
penetrated the city. , ,
- These forms are not Persian, but similar to Sogdian and Khotan baka.
If it could be proved that the first vowel offoynra was an d etymologically distm.
from d and if. Iranian would be excluded. Hut evidence is lacking. Ti etan o. rce .
baka au, Armenian o (in toxara.Uan) support o. Armenian fuxnn-l.- could be due o
an Iranian pronunciation where u and 6 were not distinguished. nt it mou < . i j ■
necessary, even if the first syllable were u. to prove the existence of u abu in the
language. , , ,
^ In the sixth centurv a.u. according to thePhoii annals (composed A.n. b.W>) based
on a report of ^ ^ yi-ei sim^ who travelled in .'.18-52-, the Hephthalites were
related to the I* report, not to the time
of the origin of the Y“'at (Hephthalites) in Dzungaria (see Herrmann, Asia M'ljor,
ii, 569). Cf. the T'ang annals. Chavannes, Documents sur Tou-Kiue ocriiievtnnx.
p. 158.
894
H. W. BAILEY —
Turkish version of the Maitreya-samiti, which are conveniently
exhibited by F. W. K. iliiller and E. Sieg, MaitrisimH inid
Tocharisch^', SBAW 1916, 414, Aryacandra {-ry’cyntry) a native
of Xagaradesa (nkrydys, probably Jalalabad on the Kabul river)
is named as the composer (yaratmis).
vaibazaki aryacindri ^ bodis(a)vt k(a)si acari
anatkak ^ tilintin toy(a)ri tilinca yaratmis
. . pr(a)tnnyaraks(i)t ® acari toy(a)ri tilintin tiirk tilinca
aqtarmis
maitri-simit nom bitikda ‘‘ maitri bodis(a)vt tuzit t(a)ngri
yirintin yirtincokii inmak ” atl(i)y onunc
iiliis nom tiikadi.
Ended is the tenth chapter called “ Descent of the Bodhisattia
Maitreya from the divine land Tusita to the World”, in the
book Maitreya-samiti. Composed by Aryacandra Bodhisattia,
the Vaibhdsika, the teacher and acdrya, in the Tocharian
language ^ out of the Indian language. Translated by the
Acdrya Prajiidraksita from the Tocharian language into Turkish.
The meaning of yarat- ‘ make, create, compose ' is happily certain.
It may be illustrated by the sentence yirig t{d)ngrig kirn yaratmis
tipcin hiltim{i)z ' we knew who created earth and heaven ’ {Khuastuanift
167-8. JRAS 1911, 291). yarat- occurs in hendiadys with it- ‘ to make ’.
as in Sdkiz Yiikmdk (SBAW 1934) 80 ktint ulus aw barqitgdli yaratyali
■ die Griindung einer Stadt und eines Dorfes oder eines Hauses und
^ "ry'cyntry with Sogdian speiling of for fi‘l (cf. Sogd. yshnt- At. skdyida-).
- In view of the statement ol Schwentner, Tftcharisrh 12, note 1, that the
Turki-'h anatKcik " I*?! noch unerklarl ", it should he jjointed out that the Turkish
'fitk'k. "nth, is quite simply cx])laine<l by reference to the isogdian "ynthw '' Indian
(llrKlielt, Du' S'jyh'L HaruhrhnftenreAtc <1. lirit. .l/w?., ii, 70, 1. 37). In Sogdian words
-‘k -’ti: -'y interchange according to the obsolescent system of nominal intlexion
(cf. Gnmi. tiotjd. ii, 7o). A form *'yntk"k is nom. to the ace. ^ynthic. It should be
vocalized ^imlukak, a derivative of indnk from himlu- (Sogdian does not preserve
(Jid Iranian h-)^ Mid.Per'^. kindny, Armen, hnduk. This i»>luk is attested in the plural
't/'Akirt induk-t {with trajected ?/*. rather than with Reichelt ‘’nom. vSg. koll. (?)")•
Turki.'-h has nicjdiried and narranircfl the v'owels, a method of adaptation attested in
other foreign word>. e.g. jrnh/i 8kt. pmdhnnn, ■'<yfny/ Skt. Pel'iot. T'oiuig-Pfio.
Ht.'U, 43!* (quoted b}- A. von Gabain. SBA IT 1!»33. had not fully understood the
Soirdian lorms. TheChine^e Pfi 4^ W y‘e/ {Life of Iliunn Tmng, Taisho ed.
203.3, p 227, col, 2, 1. 24) ajiproximates to the .Sogdian form, since it implies a reading
Sogd. prtnyh, pr'tnij", pr"tttiy’ (MuIIer-Lenz, Soghdische Texte, ii, 90-91, Reichelt
loc. cit., i. Dhuta, 99. 100. 162, 195) transcribes Skt. prajnd. It has hence passed to
Turkish. The Sogdian indicates a Central Asian pronunciation dni for Skt. Jn.
^ lor the use of -r'a ‘ in ’ (beside ' into *), cf. andfkdk tilinca ' in the Indian language ’
in the title of the Sfitra ciuoted by F. \V. K. Muller, Vigurira ii, 51 note 1.
TTAUGARA
895
Hofes Cf. also Sdkiz Yukmuk 243, ardinin yaratmis ordular ‘ palaces
constructed with jewels The meaning ‘ compose ' for yanit- was used
by F. W. K. Muller, apart from this passage, as in To^r'i inid Knisn)'.
581 : cindtkak iltdki vaihas sastar ijaratdaci ‘ der im Lande ludicns
Vaibhasa-sastras verfasst habenden (Lehrer) Further e.xamples
are to be found in the Manichean texts published by Von le Cof|. (3u
the other hand ’ translate ’ is expressed by ilwir- and aqtar-.
The phrase andtkak tilintin toy(a)ri tilinca yaratmis is not
immediately clear, ^ as the perplexities of translators testify (see
Pelliot, Tokh. 54, and note 1). Two interpretations seem possible ;
(1) the writer of the colophon wished to express two facts, (n)
Aryacandra composed the Maitreya-samiti in Sanskrit and (b)
Aryacandra. composer of this Sanskrit poem, composed also a similar
poem in toyari — he was then both composer and adapter of his own
work, and the colophon has only succeeded in expressing clearly his
character as composer, which was naturally the most important fact, —
or (2) the reference to the ' Indian language ’ may mean that
Aryacandra had used an Indian (Sanskrit or Prakrit) text as his source,
such a text as we have embodied in the Khotan Saka text (edited by
Leumann, Lehrgedicht des Buddhismus, chapter 23) or in Pali and the
Divyavadana as well as in Chinese sfitras, and out of this had made his
toyari poem. The second alternative is simpler.
Nagara, Jsagarahara, Nagaradesa, in the time of Hiuan Tsang
was subordinate to Kapisl. Unfortunately Hiuan Tsang did not record
anything of the language. A native of Nagaradesa however would
clearly have been using a foreign language in writing Sanskrit. It
is possible too that toyari was also a foreign language for him. Ve
are not however told where the Maitreya-samiti was composed. He
may have been resident in Balkh or Tarmita at the time.
5. Translations
The activity of translators in Central Asia is abundantly attested
by colophons. AVe hear of translations from Toyari, Tibetan (tirypai)
and Chinese (t’by-c tby’c) into Turkish ; from Kuchean into Toyari
and the Barcuq language (probably the language called Kanjdkl in
Arabic), a fact in no way surprising in view of the well-attested activity
of the kingdom of Kuci in the propagation of Buddhism ; also of the
translation of Indian books into Khotan Saka (from hldvdmga),
Sogdian (from ’yntknc), Tibetan, Chinese and Kuchean ; and trans-
* Schwentner’s ‘ deutlich Tocharlsch 12, is too optimistic.
896
H. W. BAILEY —
lation of Khotanese (the language of Li-yul, perhaps Saka) into Tibetan.
Elsewhere there is mention of the translation of Chinese into Sanskrit
(Chavannes BEFEO 1903, 438).
It has been necessary to clear the ground by establishing the
meaning of toyara ; and to reject the erroneous view of the script and
literature of Tokharistan, before proceeding to the complex problem of
the language known from the fragments pubHshed in Tochari'sche
Sprachreste by Sieg and Singling, 1921, and treated in their grammar
with Schulze’s collaboration : Tocharische Grajnnuitik, 1931. The
language may at first be conveniently referred to by the neutral name
(which has been used above in the footnotes) Dialect A.
Dialect A
The Maitreya-samiti of Aryacandra, undoubtedly the same poem
as is preserved in Turkish, is known in Dialect A. Fragmentary
colophons are preserved in Nos. 253a 5, 2586 3, 2596 2, 263a 6, 265a 1,
2876 3, 297a 8, 2986 4, 299a 7, 3026 6 (see Muller and Sieg, Maitrisirnit
und ,, Tocharisch SB AW 1916, 415). The various colophons allow
the following to be established :
vaibhasikyap aryacandres raritwunt^ maitreyasamitinatkam
in the Maitreya-samiti-ndtaka composed by Aryacandra the
Vaihhdsika.
Whether the Turkish has abbreviated the title to Maitreya-samiti
or the version in Dialect A has expanded the title by the addition of
ndfaka cannot be decided. The difference may reflect different
manuscript sources.
No reference is made to translation. ^ But if the Turkish colophon
quoted above is correct a Tocharian version existed and possibly,
as noted earlier, also a Sanskrit original.
In the infancy of Central Asian studies, in 1908, Sieg and Siegling,
both Indianists, published a paper treating of Dialect A and Kuchean,®
^ The meaning of the verb n/u’-, nominal derivative retwe. in Kuchean ritt- and
raitwe, is among those best attested. It translates in both dialects Skt. yog-^ yuj-
‘ to join, compose The Turkish yaratmis ‘ make, create ’ in the same context
confirms this meaning. It is impossible to justify the use of ‘ ubersetzen ’ either for
the verb ritW’ or the noun rptivr.
2 It is equally the practice of colophons in Khotan Saka to omit reference to
translation from Sanskrit, although such information may be given at the beginning
or in the body of the work.
^ “ Kuchean ” for Dialect B is now beyond dispute. Turkish hwys-n (in Sogdian
script) and kuwn (in Arabic script) is the name of Kuci {Kucha). Icusdn tili ‘ language
of Kuci ’ is conclusive. It should be remembered that kuci is the name of a country
(Hiuan Tsang used ^ ^ ^ ‘ land of Kuci ’), not only of a city as Muller and Sieg
TTAUGARA
897
entitled ‘ Tocharisch, die Sprache der Indoskythen From the existence
of the two versions ^ of the Maitreya-samiti, and the reference in the
Turkish colophon, the invalid inference was drawn that Dialect A
and toyari were the same language. It is of course clear that a version
in the language of Tokharistan (that is, in Greek script, as we have
seen) does not exclude the possibility of other versions.- At that time
however knowledge of Central Asia was vague. It was quickly
discovered that the second assumption — that it was the language of
the Indo-Scythians — was wrong. The next step was to propose to
exclude Dialect B (Kuchean) from the name " Tocharisch “, although
it is clear that if the Tochari had spoken Dialect A, Kuchean is too
closely related to be anything but a language of the Tochari. When it
became certain that toyari meant the language of Tokharistan, an
attempt was made to save the first assumption by the further
unsupported assumption that Dialect A had been imported for study
from Bactria. But there, as is now certain, the Greek script was used
for literature. Dialect A is known only in the Brahml script. We have
seen above that the script of Tokharistan was known in the Turfan
region, as the Berlin fragments attest. It is clear that a better case
for the identification of the language of these fragments in Greek
script with the toyari of the Turkish colophon could be made out,
though it is well to remember that they too may contain a still
unknown language.^ There is nothing beyond the existence of versions
of the Maitreya-samiti in toyari (according to the Turkish colophon)
and in Dialect A to justify the inference of their identity. The loan-
words in Turkish which were quoted to support this inference may be
from either Kuchean * or Dialect A. As will be seen below Turkish
seem to have imagined, see Schwentner. Tocharixch, 13-14. Sanskrit has kaucya
‘ Kuchean ’ for the people of the land of Kuci (Liiders, Weitere Beitrage zur Geschichte
und Geographie von Ostturkeatan, .S'BAU' 1930, 17). [It is very necessarj- for Central
Asian studies that all Kuchean materials should now soon be made available.]
^ The same argument would prove that Tibetan, Chinese and Khotan Saka were
identical, because the Sumukha-dharani is known in all three versions.
- It is well to remember that the author Aryacandra was from Xagaradcsa, to
the south of Tokharistan.
^ We have to remember that in 1933 near Samarkand a document in unknown
script was found. It is stated to lie written from right to left, the letters not being
joined, see Sogdiiskii Sbornik. .Academy of Sciences, Leningrad, 1934, p. 37. No. 1.').
We have also the long list of names in the colophon of the Gilgit Sanskrit ilS. (see
S. Levi, JA 1932, 1, 45 ff.), such as khiikhulhnla, khiikhuphana. utrupkann,
Urapukhra, Jerakaina. These are evidently not Turkish which does not know initial
I- (or r-).
^ A fact recognized also by Muller and Sieg, loc. cit,. SBAll' 1916, 410 note 2,
who .state that kdsp, wasampdt, pimtirnt, kaptr, ten parynn and rajngri are known
also in Kuchean,
898
H. W. BAILEY —
has other such loanwords where only the corresponding Kuchean
words, but not those of Dialect A are attested. The evidence against
the identification is positive, and the list of improbabilities involved
in the theory that Dialect A known only in Brahmi script was imported
from Tokharistan is long (the period of 600-700 years' separation of
Kuchean and Dialect A in Bactria. the different script, the significant
absence of MSS. in Dialect A in Kuci to which thev were supposed to
be imported, the long period of the independent kingdoms of Kuci and
Agni which makes an identical language in both almost unthinkable,
the silence of the Chinese as to toyari books for study in Kuci or Agni,
where they stated that Indian books were studied, the close relation-
ship of the two dialects, even in loanwords, the use of different Buddhist
technical terminology if the Kucheans had learnt from Dialect A. the
absence of positive information how a language like Dialect A in
Bactria became the language of Kuci). They are in fact insuperable.
Agxi
In referring to the kingdom of Agni ^ it is this same name Agni
which foreigners employed. It occurs in the Saka text from Murtuq
near Turfan, which is in the same dialect as the texts from Barcuq
(Maralbashi), edited by Sten Konow, Ein neuer Saka-DiaJekt, SB AW
1935, No. viii, in the form agnye gen. sing. It is known also in Chinese
in various transcriptions, brought together by Liiders, Weitere Beitrcige
24 ff. , as follows :
M 'uo-i % ian-g‘ji, jian-g‘ji
^ # '^(o-g‘ji
M fB W # M d-g‘ji-nji
One might deduce from these a native name *okni with k not g,
since g is considered to be foreign to the language of Agni, and o not
a or u.~ if the Chinese 'no and a, Skt. and Saka a are imperfect attempts
to represent the one native sound.
The Sanskrit text dealing with the domestic affairs of Agni, given
in full below, uses Agni for the country, and a painting illustrated on
^ The history of Agni (later called by the half-Turkish half-Persian name Qara-
^ahr) is given, somewhat too briefly, by S. Levi, Le Tokharien ”, JA 1933, 1, 8 ff.
It is interesting to recall that who died before a.d. 345, seems to
have been a sovereign with power extending to Krorayina (see Chavnnnes in Stein,
Anr.}pnt Khotan, 537, 513, F. W. Thomas. Artn Orievt. 1934, 49). People of Agni were
also dispersed in Kansu and the Qomul region (Pelliot, T'oung Pao, 1931, 496, and
flile«, BSOS vi, S44 ; ‘the Lung (Dragon) tribe ')
is us cl to transcribe Skt. n in wlydnn ^ 'uo -dz‘{ang~nd.
TTAUGARA
899
the title-page of Tocharische Sprachreste has an inscription which
reads ; Hsya guru dcdrya silacandra a[g]neya (ibid., introd., xii).
Agneya is ‘ Agnean, of the land of Agni The Turkish form of the
name has not yet been pointed out. It is therefore to be hoped that
it will occur in the Uighur version of the Life of Hiuan Tsang, of which
the fifth chapter has been recently edited by A. von Gabain, Die
uiguriscJien Ubersetzung der Biographic Hiian Tsang, (S5A1T 1935. ^
[A conjecture for a possible native etymology of Agni should not
be forgotten here. If *okni is adopted as the indigenous name attested
by Chinese, Saka and Sanskrit (in Sanskrit g?i is not used, hence gn
would be substituted for it), such a word *ok)li would be an adj.
derivative in -ni to *ok, as in yokaiii ‘ thirsty cf. yoke ‘ thirst ’ in
Dialect A. In turn *ok would be the form of Dialect A corresponding
to Kuchean auk ‘ serpent It is possible that this word coukh-express
both ‘ serpent ’ and ‘ dragon ’ (‘ dragon ' legends are recorded for
Kuci), but for the twelve-year animal cycle two terms were needed,
hence in Kuchean from Skt. ndga was used. In Dialect A
‘ monsters ’ are called 7idk and tmtdr, both foreign words. It is note-
worthy that in Iranian the Avestan azis dahdko, a particular azi
‘ serpent ’, supplied the later Mid. Persian azdahdy and Mid. Parthian
azdahdy as a word for ‘ dragon It is therefore interesting to notice
that the Chinese used f| li^ong ‘ dragon ’ to render the dynastic
name of the kings of Agni, and this same IjAmg to name the people
of Agni dispersed in Kansu and the region of Qomul. The Sanskrit,
as noted above, had the ethnic designation agneya. It does not seem
too rash to consider the Chinese li'^ong as a translation of *okni.
There are then two ways of understanding the development of
meaning : (1) An animal name used as name of a hero (‘ the dragon ’
‘dragon-like’), thence a tribal name (cf. daat Dahae, Dahistdn,
and Khotan Saka daha- ‘ man ’), hence to a name of the country and
also a dynastic title ; or (2) ‘ the dragon ’ as name of the king and as
dynastic title, thence used as a name for the country. The Sanskrit
and Saka a-, the Chinese d (which may bO based on the Sanskrit form)
and ’uo seem together to exclude a diphthongal pronunciation of the
first syllable, but to favour o.
It is possible that the heroic name Arjuna may have seemed in
Sanskrit a suitable substitute for *okni ‘ the dragon-hero ’, and so
have been used in the royal names Indrdrjuna and Candrdrjuna (see
* A letter from Fr. von Gabain has shown this hope to be unlikely of fulfilment.
VOL. vrn. PART 4 . .18
900
H. W. BAILEY
below). But we have also in a Kuchean document a royal name,
Ksemdrcune (Levi. Le “ Tolharien " 23, where speculations are given ;
further developed by Fukushima, On the Designation-Problem of the
so-called Tolharian language, 1935, 39). It should be noted that
-arjuna does not occur in the names of the kings of Kuci listed by
Liiders, Weitere Beitrdge, 23, to which is to be added the name of the
successor of Suvarnadeva. m m ^ ^ m xd-liei puo-siet-piet
Haripuspa (Aurousseau. T'oung Pao, 1914. 393). [This would suggest
also an interpretation of Artep as Harideva, rather than with Liiders,
loc. cit., Haradeva.]
If the conjecture f| IPong ‘ dragon ’ = *olchi (Agni) is
acceptable, a further suggestion is perhaps worth noting down. The
Chinese give also the Kuchean dynastic title. The Wei annals (quoted
by Levi. JA 1913, 2, p. 346) read : Les roi.s de Koutcha ont pour nom
de faniille Po Q . In the same paper, p. 334, Levi notes the variant
reading ^ for this Both were pronounced b‘vk. Since we have
the equation — Dynastic Title = Country's name — in the case of
Agni. the missing member of the equation — Kuchean Dynastic Title
^ = Country's name — would be ’ Kuci ’ itself. If then Q here
actually means ' white ' as has usually been assumed (the variant
since it has the same pronunciation, need not discredit this),
it would be possible to conjecture that ‘ kuci ’ meant ‘ white The
native pronunciation of the name ‘ kuci ’ would then be important.
Chinese ^ ^ kj"'i-tsi, ^ kHiot-dzH (see Pelliot, Tokh. 86 note 3)
indicate hdsi (possibly with ii), and ^ Piudt-tsie indicates kutsi.
Skt. has kuci. knclna (c = ts) and Chinese kuth may be due to Sanskrit.
Turkish kmjs’n {kiisdn) has s. Since Chinese could distinguish ts and
ts, the native pronunciation may have been rather kutsi.
An Indo-European etymology of kutsi ‘ white ' is easily found
(no proof of correctness unhappily, since one could find etymologies
for almost any combination of sounds in Indo-European) in the
base keuk (Skt. sok-, sue- Iran, sank- ‘ be bright ', which supplies
colour names in Skt. sukla- ‘ white ', Av. suxra- ‘ red ’, cf. Khotan
Saka surai ' clean '). We need not conjecture such a wealth of words
for ■ white ' in Kuchean as Skt. shows with its dharala, avaddta,
sreta. suci, sukla, arjuna, dhauta, but two words, as in Iranian, Av.
aurusa- and spaetita-. spiti-, Mid. Pers. arus, spc8. would not be
excessive, so that (at least earlier) Kuchean mav have had a word
kutsi ' white beside arkud ’ white '. A word kutsi could be considered
an adjectival derivative in -i (cf. arkwi, IE. *argu-, poijsi ‘ sarv'ajna ,
TTAUGARA
901
Dialect A wsi ‘ yellow ' : u'iis ‘ gold ') to a word ' brightness,
whiteness', from IE. or *kulci-, cf. Skt. suci-, Av. sxka-. sxca-.
In Kuchean and Dialect A k is palatalized to s. presumably through
a stage is (cf. is > s). In certain cases we have A’ > c ; after a nasal
in cuk- ciix- enis- ‘ take ’ in all three stages, kxtsi would then be an
earlier form of *kusi. Such a *k)isi could be the source of Turkish
kiis- (in kwys-n, k>i:s-n) with s, although here the development may
be is > .s.
If this conjecture should be confirmed, it would not be desirable
to seek arjuna in this It would avoid the difficulty that arjxxa
is attested also in the names of Agnean kings where 1^3 has no place.]
The fragments in Dialect A were found (always associated with the
closely related dialect of Kuci) in the kingdom of Agni and in the
Turfan region. Once the erroneous confusion with toyari is forgotten, it
is clear that there is a slight presumption that the place of discovery
may he for these MSS. the place of origin.^ 3Iost of the MSS. of
Dialect A are literary,^ largely if not wholly translations. But one MS.,
Xo. 370. has a more prosaic purpose. The description given of it in
TocJiarische Sprachreste is inexact, and written at a time when the
full document was not known. It was only with the publication of
the whole by Luders, Weitere Beitrcige, 24-5, that the character of
the document could be realized. It is necessary to quote it in full :
Recto
1 . tad-artham avasanibodhayami yad ayani mahatma aparimita-
subha-rucira-punya-pra nwhaddnapati agnisvara agni-
maharaja indrarjunena sardhan; agni-maharaj/li-
2 . ya survaprabhava sardhani sarvai pahca-gati-paryapaniieh satvair
yo^sau bhagavac-chravaka-saingham anena varna-gandha-
rasopetena aharen^opanimanitrampayati tasMcld d-
3. hara-pradanad punyam pnnyabhisyandah yas ca kusalam ku-
salabhisyanda tad bhavatv etcsain dayaka-danapatl/m»i drste
va dharme ayur-varna-bala-sukha-6/mgaisra-
4. rya-paksa-parivarabhivTddhaye_stu idas ca teya-dharma-pari-
^ It IS almost amusinrr that llio (Hscovcry of Kiichoan Texts in the Kincrdora of
Almu (tln*v Mf'FC found also in the Turfan roition ami in Tun-huang, Stein, S>rindin,
II. !‘l."' ha" heon u.^ed as an arcrument that Km-hean wa-< also the indigenous language
of the uftf-n ho";tile ALUiean'J.
- Tof hfin'-rhp Sjn-m'hr- d* , jntn»d. v. " uut io HupIk rn Pelliot seein-i to have crone
farther, '/'ckh. 63 : (/'/ant af/ " dinhrft .1 ", // ji'td qne pur /< '' ihutuKrrits
'Vun cnrarf^re litlerulre.
902
H. W. BAILEY
tyagat maitreyanam sarvesam bodiii-inarga-pratipamiianam
ksiprdbhijnay^astu tatha brahma-sakradlnam catu-
5. rnam ca lokadhipatinam astavisatis ca gandbarva-kubhanda-
naga-yaksa-senadhipatinam prabhavabhivrddhaye^stu : tatha
agiii-\dsaya-paripalakanam devatanam vyagra-ska-
Verso
1. ndbaksa-kapila-manibhadra-prabhavabhivrddhaye^stu : tatha
kumbhadhipatinam srisambhava - lohitabha - krhisa - svastika -
-indra-prabhrtinam prabhavabhivrddhaye^stu : tatha ndgadhi-
2. patinam maiiivarma-sudarsana-susukhah prabhavabhlvrddhaye^
stu : tatha purna-agnindranam adau candrarjunasy^abhyatl-
takalagatasya upapattlvisesatayai-
3. r bhavatu samasatah panca-gati-paryapamnanam satvanam
caturnahara-parijnayair bhavatu yac ca kinci dyate tat
sarvebhya samasamo datavyam iti : — ||
4. paklyossu pis-sahk-si Semi naktas naivasikassi tarmatam ayis —
kus ne naktaS naivasikan trinemintu passi wrapos — bram-nat
w’la-nnat stwar idwe na
5. kcini las visnu mahisvar skandhakumarass aci — viki okat pi
tassi nan yaksan kumpantan kintaren kandharvin tkarn-sini
eppre-sini kus pat nu naktan
The Sanskrit, although not strictly grammatical, will be under-
stood (it is paraphrased by Luders, loc. cit.), but it will be well to add
a translation ^ of the passage in Dialect A :
Let it (i.e. the Samgha) hear. May the Jewel of the Bhiksu-
samgha give the dharma-ddna to the gods and Naivdsikas,^ %vhatever
gods and Naivdsikas have deigned to guard the triratna, the god
Brahma, the king-god {= Indra), the four great divine kings, Visnu,
Mahesvara, Skandhakumdra, the 28 leaders, the ndgas, the yaksas,
the kumbhdndas, the kinnaras, the gandharvas, whatever earthly or
aerial gods.
The importance of the Sanskrit portion of this document was
recognized by Luders, but it is equally important for Dialect A. Its
^ After Siegling, apud Luders, loc. cit., 26, where, how’ever. probably bv an over-
sight, stands “ Skandha, Kuraara ” as if they were two different gods. Khotan Saka
uses skamndhd aysamnai^ Skt. skanda-kumdra.
The naiidsika is known also in Turkish. Muller, Viguricoy ii, 83, has nitvasiki,
^80, naivaziki ; in the Vighur-Chinese glossary- n^yb^syky n-aiwasiki is explained by
good genius The Mahavyutpatti has naivdsika explained as ‘ inhabitant
TTAUGARA
903
character is e^’ident. It is concerned with a gift (dharma-dana) by
Indrarjuna king of Agni to the Buddhist community of Bhiksus,
and they in turn are exhorted ^ to present this dharma-dana to the
protecting gods, not only those already invoked in the Sanskrit part,
but others also, so that the offering is now made universal in the
portion in Dialect A. It is clear that both parts belong closely together.
According to Liiders they are both written by the one scribe. Similar
Sanskrit documents are published in this same place by Liiders
concerning the kingdom of Kuci. This present document is however
peculiar in its use of Sanskrit followed by a passage in Dialect A. To
an unprejudiced reader no more satisfactory evidence could be had
that Dialect A is the language of the kingdom of Agni. It is certain
that the document was -written in Agni. The sacred language is
Sanskrit.^ It is likely that the indigenous language would be the second
language. This is not a literary document imported for study but
a domestic concern of the king and the samgha of Agni.
But there is more. That the language was not a static dead language
(such as a foreign ‘ sacred ' language must be. as in the case of Sanskrit)
is significantly observable in precisely this document. The forms
nan ' nagas ' compared with the nakdh nagdh of other texts (Kuchean
sing, ndk ) ; bram-hdt ‘ the god Brahma ' and wla-hnat ' the king god
with Mt, hndt for older nkdt (Ktichean hakte) ' god ' ; Ids ' kings ’
for older Ians •, and the isolated wrdpos beside -ivdrpo eight times and
six times u'drpords, are signs of phonetic change, of developing
language. Similar developments can be traced in the documents of
Khotan Saka, where the older religious texts show a far more archaic
language than the secular documents, and indeed than the later
religious texts.
We therefore assist here at a living changing language. It is
not merelv uncertain orthography (as Schulze's remark DLZ. 1923, 47,
“ zeigt in seiner Ortkographie erne im gauze bemerkenswert sichere and
gleichforrnige Haltung ’’ might imply), but a later stage of the language,
a fact of great importance for the study of Dialect A.
If this fact is fidly recognized it will be found that all indications
^ The e.xhortation (paklyossu = “let it hear”) is on the model of the Sanskrit
rubric inwtv dryasamghah.
- No. 414 contains prescriptions for the posatha-pravdrand of the bhikmnis
(nuns). The formulae to be uttered are in Sanskrit, but the instructions are in
Dialect A. Here too we shall see the indigenous language used to explain the sacred
language. No evidence exists nor is there probability to make credible the existence
of two sacred languages in this one conntiy.
904
H. W. BAILEY —
fit perfectly. Dialect A is the Agnean (Skt. Agyieya) language. The MSS.
belong to Agni where many of them were found. The language is
closely similar to the language of Kuci, but not identical, as is natural
in the case of two independent kingdoms, separated by mountains and
difficult roads, whose independent history can be followed from at
least 102 b.c. when a Kingdom of Kuci is mentioned. Hiuan Tsang
clearly considered the literature and languages of Agni and Kuci to
be distinct but he remarked that Bharuka had a language similar to
that of Kuci. Kumarajiva (horn in 344) translated tulchara by /]> ^ K
who have no attested association with Agni or Kuci, therefore though
a native of Kuci not acknowledging that the name tukJidra applied
to either Kuci or Agni. Hiuan Tsang knew tuoxudld in the west, and
heard of ruins in the east on the southern route, but he does not
associate the name with the cities of Agni or Kuci on the northern
route. There is no evidence to prove that Dialect A was not spoken in
Agni. MSS. of Dialect A were not found in Kuci (an argionentuttt, ex
silentio, but in this case of importance, since if Dialect A were imported,
Kuci lay on the route ; but if Dialect A is Agnean, the finding of MSS.
of Dialect A in Kuci would be indifferent). Kuci was eminent in
Buddhist studies, Kucheans, immigrants as it seems into the kingdom
of Agni and the Turfan region (where in Sangim near Turfan, like the
Eussians with their inscriptions in the church in the rue Daru. Paris,
and the English with their inscriptions in the many English churches
outside England, they too wrote their inscriptions on the walls of
their shrines) were clearly as interested in Agnean MSS., as their glosses
in Kuchean show, as were the Turks, whose Turkish glosses are
preserved in No. 394. A further decisive proof that Dialect A is a
language of Central Asia, that is, of Agni, is furnished by the loanwords
in Kuchean and Agnean.
The name Agnean ^ will be used in what follows for Dialect A.
If it is necessary, it will be convenient to use Agni-Kuchean - as a
name for the earlier form of the language whence are derived the two
dialects of Agni and Kuci, including the language of Bharuka (/c.v^t’
Hiuan Tsang) and the possible traces in the language of Krorayina
found in the Niya Kharosthi documents (Burrow, JRAS 193-5.
667 ff. ; cf. Liiders, BSOS viii (1936), 647).
^ The name laracharien proposed by Levi, JA 1913, 2, 380, and adapte.I to
Karashakrian by ^Iironow, Borz. Orient. 6 (1928), 89 fF., is taken from too late a period
to be acceptable. For Agnean we have the warrant oi^kt.agneyn and the contemporary
name of the country’ itself.
* On the model of Indo-Iranian.
TTAUUARA
905
Loamcords ^
1. sosta^axL is found in 222 a 2 ;
lane amasan sostankan sne-paltikan praksantan
kings ministers offdals merciless - exactors
TochariscJie Gram., p. 106, oifers no translation of this word. It
clearly represents an older form of the word in the third cen/tury
Niva Kharosthi documents .svathamgha, sothamgha, sothaniga. sothdga,
an official title. ^ So far the word is known only here and in the Xiya
documents. The consonant group .st is used for Indian s/ in ucchist
and occurs in several indigenous words. It occurs also in Barcuq
Saka, but not in Khotan Saka, where -st is written. The word may be
an indigenous Krorayina word, but there is a likelihood that official
titles should be borrowed.
2. katak, Kuchean kattdke ‘ householder ’, as a technical Buddhist
term. The word corresponds to Khotan Saka ggdthaa- which together
with Sogd. hrtk. brt-k is derived from a Prakrit form of Skt. grhastha
(Hansen, BSOS viii (1936), 579-580). The Saka could be the
immediate source of the word in Agnean and Kuchean.*
3. arsi.
In 251b, a passage of the Maitreya-avaddna-vydkaraiia, occurs
the following passage,® verse 4 ;
kulmass or e- sokyakal tane maheam s- -i <6 syllables'} :
<7 syllables + metrak-sinaip) opslyassal syak kumnassi ;
sokyokal ® niim arsissi kasu taki<s-> . -c kasal -i <1 syll.} :
<1 syll. + sokya>ki’d sraddhessi ma tiri nas miintat.si kar-
samantap :
^ Indian loanwords in Kuchean (Mironow. Kuchean Studies^ i, Bocz. 1928 ;
WinAiwT, Snwikrit drugs in Kuchean. JKAS 1925) and in Agnean {Tocharische
' rranonatik, have already attracted a large amount of attention. Other
loanwords have received occasional notice (bibliography in Sf-hwentner, Tochan.cch 46).
“ Cf. the passage, G4 b 2 ;
sne-karnm sam lyalypu hkat
sne-paltikah cem haktah ku.'i ne cami
The context of 222 a 2 (dcs<Tiption of an evil period of time) makes it likely that
here cine-pdltik may Ije jiarallcl to ■'<nf-kdnun ‘ mereiiess In form is, according
to Toch. Gram. p. 13, ‘ ganz unklar
^ In Krorayina the sathainijha was a tax-collector,
^ Toch. Gram. p. 13, curiously compares XPers. kad-xndd (so to read). MidPers.
uses Itkhivby kaoay-x^'cihny ‘ master of the house, governor of a province \ but the
second component is indispensable.
' The contexts are not so colourless as they teemed to Levi, Le “ Tokharien ”
( . 6.
'■ The variant in 2olb has sokyakal.
906
H. W. BAILEY —
Translation :
.... greatly is desire thus ....
... to come together with Maitreya’s consecration.
Greatly is it the desire of them, the Aryas. Good may it be .. .
together . . .
. . greatly is it the desire of the Sraddhas {believers). It is not
the way of the Prdjna {ivise man) to be passionate.
The parallelism of drsissi and sraddhessi beside karsdmant- (= Skt.
prdjna) in a poem of the Maitreya literature assures the meaning drya
and srdddha, both, as well as prdjna, being Buddhist technical terms.
drsi < Central Asian Prakrit *drsa- Skt. drya.
In Central Asia beside the Sanskrit attested in Sanskrit texts and
in numerous loanwords in Agnean, Kuchean, Khotan Saka, Sogdian,
Turkish and Chinese, a Prakrit is found in the Dharmapada MS.
(MS. Dutreuil de Rhins), the Niva, Saca and Krorayina documents,
and loanwords in Agnean, Kuchean, Khotan Saka, Sogdian, Turkish
and Chinese.
In Khotan Saka these Prakrit words are easily to be distinguished
from the literary Sanskrit words. They further illustrate phonetic
changes which differ from the changes undergone by Iranian words.
It is therefore possible to know the forms of the Prakrit whence they
came. So e.g., Iranian i- is in Khotan Saka, as juvdre ‘ they fight
base yaud-, but 3ama is ‘ the god Yama ’, showing Prakrit s- < i-.~
The consonant group ry appears in this Prakrit as rs : in Khotan Saka
virsa, Skt. idrya ; ttdrsasuni, Skt, tiryagyoni. In the Niya documents
ry has been regularly used in the transcription, except in the one
uncertain case 572 (covering tablet, reverse) ; suryadade or sursadade.
Graphically it is clear that a decision between ry and rs would be
difficult. If rs is correct, even in this one case, it would be necessary
to assume that in rs unvoiced s had replaced the expected z (however
written) of rz, just as k t p so often replace g d b respectively (Burrow,
JR AS 1935, 667 ff.). If ry is right, it could, if necessary, be under-
stood as retention of an historical spelling. But the forms with rs
For one would expect in other texts a spelling *7,rja, indicating *7irza.
Cf. (if they are not due to literary pronunciation of Sanskrit) A1 Bairuni's j\
’rjbhd and Abu 'l-Qksim SiVid b. Ahmad b. Said’s j| <rjbr for Aryabhata quoted
by Gabriel Ferrand, BSOS vi (1931), 336, note 4.
* Single i expresses c in the older Khotan Saka.
TTAUGAEA
907
in Khotan Saka suffice to show a Central Asian Prakrit with the change
ry >
It is evidently this same Prakrit which has given drsi to Xgnesin.
The final -i may indicate that the word had passed through Saka
(nom. sing, -d, -i), but this cannot be insisted upon.
It is well-known that in Central Asia Buddhist technical terms
were often translated, whereby the indigenous word was given the
full Buddhist meaning. But it was also a common practice to take
over the Indian (Sanskrit or Prakrit) term into the language. In
particular Sanskrit literary compounds are abundant, though they are
rarely of linguistic interest. At times both methods were adopted,
so that a double (or, if both Prakrit and Sanskrit were taken, a triple)
vocabulary resulted. The following selection will suffice to illustrate
this :
Khotan Saka.arahanda
asana-vajsama
Skt. arbant
aryastagamarglnai
pajsamana asapa-
‘ worthy of honour ’
hasta padya beysufia
aryastahgamarga
adj.
pade
Chinese Si
M siang
arya
^-Iji-ia (and passim)
Kuchean. Agnean
Kuchean passenca
payattika, payitti
payti
{Toch. Gram.,
Kuchean arhante
asanike
61, note 1)
arbant
Agnean arant
asanik
arhant
karam
lyalypu
karma
tarm (dharm)
pal, niiirkanipal
dharma
abhisek
opsaly
abhiseka
To this is now to be added :
Agnean drsi
klyom
Skt. dry a
beside the literary dryamdrg- Skt. dryamdrga-.
In other contexts arsi is not so clearly defined.
294 a 6, a passage of the Maitreya-avaddna-nataka, is merely a
fragment, but by its mention of the three jewels {tri neviintu), the
King A^aisravana {vaisravam Idnt) the lokapdla of the north, and the
’ In Khotan Saka itself rs can indicate rz. If the Agnean word were direct from a
Prakrit rz th.e ch.ang6 to would, be due to the same tendency in Agnean, as in the
Niya documents, to replace voiced by unvoiced consonants.
908
H. W. BAILEY
names of the disciples <Kan>ndinya, Asvajit and Bhadrika, it is
proved to be an integral part of the poem :
. . . opsaly matny arsissf tmassal ....
. . . {Maitreya’ s) consecration, s&t^ot of the Aryas therexrith . . . .
arsi ype, occurring on a fragment without content- is clearly Skt.
arya-desa. Tibetan uses hpkags-yul = drya-desa for ‘ India
a phrase natural in a poem of Maitreya.
383 ala metrical text :
: tiim surma si arsissi kalka-m nom kl’
for this reason he icent for the renoiim (? nom klyu) of the
Aryas ....
More interest attaches to the three other passages where arsi
is used, in Nos. 229, 230, and 236, all in the Maitreya-avaddna-
vydharana. All three passages are autobiographical. It is probable
that a Sanskrit original existed, but unfortunately neither this
presumed original nor the author's name has been traced. The passages,
229 and 230, refer to the composition of a kdvya (230 b 2 ydniit kdvvi
‘ you are enjoined to make a kavya y an injunction from someone
styled ‘ great king ‘ {tsopats n'dl) whether human or divine, which the
poet proceeds to carry out).
^ Skt. aryade^a is doubtlofs ‘Central India’ in Rdjaiarangmi, i, 315, vhere
it w used in reference to the raids of Mihirakula, and the adj. aryaddya. ihid. vi.
89, where a college {malha) for students from Ary.adcia is mentioned. In tlie pro\erb
(Bohtlingk, Indische Spruche 1025) dryadeia-kiila means ‘ a family of India ’, where
Bbhtlingk rendered literally ’’ im Lan-de iter Arjn ", The Tibetan titles of two medical
texts in the Tanjur contain the words hphags-yul, which Cordier rendered by
aryndeia (Calcd'igiie du fundi liljitnin, iii, ji. .502). Mdo-hgrel 151. Xo. 5. was composed
by hphags-yul phfitahahi sman-pn danndnia ' the jihysici.in D.Inadasa of Phataha m
Aryadesa ’, which is explained by the st.atement that Phataha is a place {yul-gru)
in rgya-gar dbus-hgyur ‘ the central part of India ’ ; Xo. 7 was written by hptnigi-yul
dhu-i-hgyur mathinahi rgyad-rigs kyi smnti-pn mghundtha ' the physician Raghunatha,
a Kijatriya, of Ilathnra in the central jiart c>f Aryadesa [Cordier iin es in the Cntalogue
Magadha for dbus-hgyur, although he had earlier, BKFLO 1903. 628. rendered by
' rinde centrale {rgya-gar dbus)’ and 'le midrein Raghunatha. de Mathura, dans
VInde centrale ’. If ilag.adha is right, the Tibetan knowledge of Indian geography is
inexact.] hphagi-yid is here equated with rgya-gar ' India ’. In a letter of 19.12.1935,
Professor F. W. Thomas informed me that hphags-pahi yul occurs in the Ladakh
Rgyal-rahs, ed. A. 11. Fraiicke. p. 25, 1. 3 ; and in the sub-title of the Jjbag-hsam-ljon-
bzais, w'hose author was born ni .a.r>. 1702. ed. S, C. Das. p. 1, we have hphagi-yul
rgya-nag had ' India China Tibet ’. lYe shall probably prefer to render hphagi-yul
by aryadda rather than by the drydrarta adopted by S. C. Das in his index.
Professor Thomas has also pointed out that in the dictionary ihe-rin-dban-rgyal,
hphags-yul is rendered by sii(read : pu)nya-desa and madhyadesa .
TTAUGAHA
909
It is necessary to give the whole of the very fragmentarv text.
The author is shown (229 a-229 b 6) meditating on karma, niraya
(hell), and the Mesas, which leads him to contemplate the composition
of a kavya on Maitreya’s wonderful deeds. His expression of this
intention is of interest here.^
229 b and 230. Metre 12 + 15 + 12 + 15. Verses 52-64.
52. ] . tain yarmam tarn hi cam ka\"\u-si retweyac :
2. ime piiltsak yes arsi kantwa ritwassi kanasal ;
3 sk- tak pam kavvi-si Tehveyam ....
4.
53. 1. . ke arthantu puk anemsi ritwaslani :
2. pasiil lyalypiiras ca . . . .
3 kam ma pastiir arthii ....
4 wwriampal skarn sasparku maskatranr :
54. 1. cincar ma wenas klyosamsantah martanta ;
2 -s katkenc a
3 praski sance taka hi ;
4. sarki saheentu mak katkar hi rakentu maskaskla :
55. 1. tamyo md campu wiirtsyo pakar ritwassi :
2 -t p- . lo :
3. markampai-sini nu imeyo tiis raritwa :
4. skaya arth passi lakentu nu maskaut hi mak wiikna ;
56. 1 tarn ;
2. kraheiin markampal sparkaslune mar hi tsani/uTntsu tas :
3. paltsankani nunak kavvintwassi kalymeya/p :
4. -m
o7. 1. . . s- wawo papsu sparkfdune vis ;
2. t)na >arki rakodic . heat liihci mamaskunt fdkont pe ;
3. cam tarya sarki skayct
4 //Kirkampal :
5S. 1. cesmi mosaa/i ats rarituvl !(«km ats ka^■vi :
2. miint -ni -e . -i -ii metrak-sinam opslyis tarn kalymam cam :
3
4. . . . takehc m- . kiim- . . -he ;
59. 1. a . w- . tsopa . . . . -m hass arsat :
^ Sieg translated the passage in SBA It' lOlS. v. infra, but the context had been
misunderstood. It is defined clearly by the fragmentary conversation with the Great
King (‘ Maharaja '). The poet has lieen enjoined to make a Kavya. All must therefore
accord with that fact. Sieg's version was distorted by mistranslation of ritxo- and
retwe, and his unsupported interpretation of drii.
910
H. W. BAILEY — ■
2. yamit kawi cam <u ma ka- . . .
3 -la ma tanksal :
4. yamtse tiis onani nunak . . . mu skat ni ma cimsa :
60. 1. prasku lyalypuras mar markampaZ . . .
2 tsopats wiil nunak :
3. raritwa kavvi kos ne ime ialka ni :
4. kanis ayatica kuc ne ma tam yomu tsam artlianZw :
61. 1 kavvi . ;
2. Eusanti takis ni mant ne macar mkaltont se kusanti yas :
3. talo akna<s . . . . -k aklu markampal :
4 nu :
62. 1. kuc ne kranc wrasan karunikan knanmanas :
2. mak mank trislune . . . . -m etstsantar :
3. tsru y-
4 kw-nnenc cam puk mankant tarnencam :
63. 1. kuc ne nu yomu taki? arthawZwam
2 takis kaka- ....
3 . -1 ;
4. tampe kalpitar kulis tampe omaskern klesassi :
64. 1. a-
Translation :
(52) In this icay therefore for this composition of a kdvya, my
thought and opinion went to the composing in the Arya language in
metre was clear in the composition of kdvyas .... (53) . . . all
meanings to be well put together. Karma nmst he ^vatched .... the meaning
is not watched .... the Dharma also is transgressed. (54) He does not
speak finely, the hearers love it not .... they will arise Fear
and doubt have come to me. Care and doubts in great number arose in
me for the lack of tvords. (55) Therefore, not being able to compose at
great length . ... I have composed this with thoughts on the Dharma.
I have striven to watch the meaning, but words largely failed me. (56)
.... May I not have sinned against the good Dharma ... 7 think however
in the nmnner of kdvyas (57) . . . guided, watched, error comes
accordingly words failed, others also. After these three
1 have striven the Dharma. (58) Because of these, I have
composed the ... . kdvya, that I may share in the consecration of Maitreya
they will be ... . (59) “ . . . . great .... caused a longing. May
you make the kdvya is not to be hindered.” I made accordingly
. ... it followed (1) tne. I could not. ( 60 ) Through fear of Karma may
not the Dharma however . ... O great king, I have composed the kdvya.
TTAUGABA
911
as far as thought came to me. In suiting the metre, if 1 have not here
attained to the meanings. ( 61 ) .... the Jcavya. Indulgence he mine as
a mother indulges her little son. Miserable, ignorant .... learnt the
Dharma ( 62 ) If good beings, merciful and understanding . find
in it many faults, few nmy they overlook (!) and forgive all
these faults. ( 63 ) But if the meaning has been attained . ... it should be
.... Let it be deemed the 'power of .. ., may the evil power of the Klesas
perish (?).^
In 236 also there is reference to the klesas {klesd^i ndkdm ' blame
of the klesas '). Then follows (much being obscure) :
236 a 7. kalpintar skuntu tsalpinc ....
236 hi tsanak sastrantu : 7
arsi nu kantwa ses kar ne . . .
. . . puk retwe yalyminn atsarn tam some puk sarsar . . .
. . . sne m- knanenc yalymi pal kantu ....
. . . ayatwa nes sarki kupre maskantar ni . . .
. . . knanmune wa . . -s ma pe sastra tsopatsam . . .
... -1 sastra nas kanam ma penu . . . . r-e sni kantwa-si
retwe . . .
. . . Ito pat kuss atsarn ne avikar -i nasmi wr . . t . -m .
yamu :
Translation : —
7. 31ay they find happiness (sukha), may they be delivered ....
. . . here the sdstras.
8. Because (I ) in Ary a language is ... .
■ . . every composition ....
. . . they knoiv . . . Dharma language . . .
■ . . in suiting {the metre) sooner or later if I have failed . .
. . . knowledge .... nor in the ivhole sastra . . .
■ ■ . the sastra is in metre, and also not ... a composition of one’s
own language ....
Sufficient can be translated to show the similarity of the context
to that of 229-230 above. It is here specifically versified sdstras which
are considered. The essential phrases, however, are the reference to
arsi nu kantwd and shi kdntwa-si retwe. The ‘ Dharma language ' is
no doubt a reference to Sanskrit.
* Or read kulis as gen. sing., and translate ‘ Let it be deemed the power of
(kulis), the evil power of the Kle^s ’.
912
H. W. BAILEY —
It is clear then that the author composed his kavya (ritwdssi
kanakil ' to compose in verse ') in the Arya language [ark kdntira).
which was not his own language (236 b 6 ma penu . . . r-e sni kantwa-si
retire ' also not ... a composition in one's own language '). If the
original text was Sanskrit, the drija language (like the hphags-ska'}
= drya-bhdsd of Tibetan) ^ will be the language of arya-desa. that is,
precisely the Sanskrit. It is then evident that the author wished to
express his sincere if perhaps too humble feeling that he lacked
proficiency in Sanskrit composition. This might mean an acarya of
some Buddhist country outsiflc India.
It is interesting grammatically to notice that drsi is here used as
an uninflected adjective (cf. Toch. Gram., pp. 251 ff., ‘ Irdeklinahih
Adjektiva ') in ark kdidird and arsi iiu kantv'd. It cannot then be
compared directly with the separation of the compound in prattika
cam patlamhkat {Toch. Gram., p. 250).
The word nr.k occurs also in the name of two metres : drsi-lancinam
adj. loc. sg. ■ belonging to the arya-raja and cirsi imkramuntam ‘ in
the metre niskramant of the Arya(s) '. It is perhaps worth while
recalling that Sanskrit also knows (a very different) metre called dryd.
It is difficult to see how in these contexts Sieg could have imagined
a native name for Dialect A {Ein einheimiacher Name fur ToxG.
SBAW 1918).- It is to be feared that regardless of context he jumped
to this conclusion because at that time when Central Asia of the
sixth century a.d. was still a nebulous land, he remembered the Greek
''Aaioi and the Latin reges tocharorum asiani of Justinus.^
The difficulties, historical and linguistic. which this has caused
^ nnja-bhn'i't ’Sanskrit’ is attested in the seventh century a.d. outside India,
corresponding to Chinese in tho colophon of the Sanskrit-
Chine-rie dictionary of ^ (Turkish kytAij, I-Tsing, born 63o),
entitkMl ^ ‘ Hook of a t!iou.sand Sanskrit characters ’. The colophon
roads : cina ak<ani sufui-sut maJo art/a hha^a stnnpta that is probably : aryahhiiAa-
r.hLakytra-'<(iha'frninnla ■saniapta [or snhn.^ramfdo' tmma pudaknh ■^amnjdah]. See
Ragchi, Dfiix hxlqurs 11)29, pp, 217-lHand 330 ( = " Taisho ed. 2133).
2 Involvinc^ also a violent mi.>rendering of )(tw- and rrfwr.
^ Speculation.^ on these ‘'Aaiot A'<i(ntt are at present of little use. We cannot
be sure from the Greek and T-ditin texts whether the At<iani were kinss of the Tochari
before or after their .settlement in Tc’kharistan. The name seems to have a suffix
-ana-, which is familiar m formin" adjectives in Iranian. It might mean that the
were Iranian or that it i.s a name bestowed by Iranians or a name which had
reached the Greeks through Iranian channels. (Theories are offered by Charpentier,
ZDMG 71, 347 ff.) '
^ The most recent attempt linguistically by X. Fukushima, On the Designation-
Problem of the so-called Tokharian Langungey Memorial volume dedicated to Katsuji
Fujioka. 1935.
TTArCARA
913
subsequent investigators are notorious. But the second century b.c.
is not the same as the sixth century a.d. in Central Asia as is now
gradually becoming clear. Dialect A is Agnean without attested
connection with the Tochari. It is regrettable that no one thought
to examine into the basis of this unfortunate inference.
4. Kuchean tvaiikarai obi., (vdtilaro nom., fvdnkaracce adj., is
Khotan Saka ftumgare ‘ ginger ’. It seems that the word was borrowed
as irdiihami, but the -ai caused it to be associated with the obi. case
in -ai. Hence a noni. in -o was created on the model of obi. p'il.sal-ai :
nom. irlt.sako ‘ root ’. Iranian -u- is here replaced by -id-, as by -ira-
and -irci- in the following words.
o. Agnean. Kuchean kuncit, Kuchean adj. kuricltu^se. hu'anrlld^^e.
■ sesame '. Old Iran. *kunx‘ita-, Mid. Pers. kwnryt, EBaloci kmiclO.
Armen. kncit‘, Khotan Saka kiimj-iala. The form represents an ohler
stage than Saka kmnjsata.'^
6. Kuchean kurkamassi beside kinirkamd.^.d adj. ’ of saffron '.
Khotan Saka kurkum, Sogd. kwrkwnph, ilid. Pers. kurkion. Turk.
kiirkdiii (Rachmati, Zur Heilkiitide der Uiijuren. SB.! 11' 1930. 19).
7. Kuchean ahkwai ’ asa foetida Khotan Saka amgufdd.
Armen, (from Mid. Parth.) anginai. NPers. anguzad. Chinese -
ra II -d -»gj'''ei and [fi -iang-g'j" i. From the Kuchean (or an
unattested Agnean form) comes the Turkish ^nk-pws {dngdhiis.
aiikubiis (?), Rachmati. loc. cit., p. 16). Saka -u- is replaced by -ica-
and -sd- by -.y®
8. Kuchean kwaiam, kiconH-. Skt. gidma, Khotan Saka gaumd.
^ Similarly a stasje of Iranian older than the Khotan Saka texts is attested by the
Xiyu Kharosthi l)ociinieiits, where rodotia ' niudder contrasts with K-hotan Sak<i
rriinai - , *i(iudanakfi-,^ WBiiluci rodan (Bvirrow, vii (1935), 7S7). In the case
of Khutnnn. the change can he seen in ol.ler ^>aka kraMna, later Saka htfunna. The
Bareuq Saka texts have hunpifui t>. 0, and hradmi 8 h 0. Both words probably refer
directly to the inhabitants of Khotan. I cannot snppoM- with Konow {Bin unter
Sfda-Duihkt, pp. 30-31) that the writers were using the word of themselves. Tlie
contexts are not (piite clear, but this view at least scenis excluded. 1 suspect tliai
k'nh< <ikf. f>f S a 3 IS connected witli the name kdnjfiki (the word is attested as kurk-
and khik-, implying at least a Turkish pronunciation kdtt,ak) applied to the non-
Turkish ])copic near Kasyur. Kasyan (i. 31. lines 5-*)) remarks : ua U-lSBytir m^rttiqu
yvtiihdhiwu flhd hi dkatijakiyydti ‘and the districts (ni-daq) belonging to Kasyar
speak in Kanjaki I am indebted for the referenee to ^Minctr'ky. We should
prohahlv recognize the same name in the Turkish k'ttc'k *k(incfik of the L ighur
do( uuient published by Haneda. Toyo Bunko, Mfmoirs vi ( 1932). p. 3. 1. 2.
- Lanfer. Bino-Iratuai 361. who hud first seen the connection betw een the KuchcTii
and the Chinese words, quoted the Kuchean without the final -s.
' According to ^ Q yd^'ei-iizdt, born a.d. 680. quoted by Levi. JA 1915. i, 89 ;
C est seulement en arrivant dans le region de yu-Vifn (Khotan) qu on en voit.
914
H. VV. BAILEY —
gom-a. Kuchean k- replaces Skt. g (cf. kattdke above No. 3) with -wa-
for -M-, and r for 1} Meillet's hesitating comparison with Gr. /Son/Stac.
Skt. gavint is probably less acceptable {JA 1911, 1, 453).
9. Kuchean aiir^, arirdk ‘ myrobalan Khotan Saka halirai
< *haliraka-, Mid. Pers. hlylk NPers. Jiallla, ihllla], Skt. harltakl.
If r has here replaced I (as in gulm-a above, No. 8), it is probable that
older Saka is the immediate source. Hence or from an unattested
Agnean form, the Tiu'kish arir-i (Eachmati, loc. cit., p. 21).
10. Agnean mahii!^ plur., ‘ buffaloes ’, Khotan Saka mahairslna
adj. ‘ of buffaloes Skt. mahisa. It is probable that the Central
Asian Prakrit had rs in this word, or that the word came from Saka
to Agnean. Tnner-Agnean development of s > rs under the influence of
kayiirs ‘ Stier ’ (so Tock. Gram., p. 107) implies too great an isolation
of Agnean.
11. Agnean a$am, Kuchean a.sdm ‘ worthy ’ from Khotan Saka
dsana- ‘ worthy ’ : dsana- < *dlmna- < *arxsana-, an -s- derivative
of arg- ‘ be worth ’, Skt. argJi-, arhati, as Khotan Saka dlsta ‘ ripe '
participle to daxs-, -s- derivative of dag- ‘ to bum, to ripen ’, and Av.,
Mid Pers. baxs- ‘ to bestow ’ to bag ‘ give a share ’, Saka J^tmhe (Tib.
bde-legs ‘ welfare '), bnmam, baimana, b(nmand, ^mand < *bax^-
man-, Mid. Pers., NPers. baxt ‘ fortune
12. Agnean aijanik, Kuchean asanike ‘ arhant derivative in
-ik from dsdrn, asdrn, No. 11. Khotan Saka uses dsana-vajsama
‘ worthily honoured ’ and pajsamdnd dsana- ‘ worthy of honour ’
in this sense.
13. Agnean mahor, Kuchean tnahur, nmhur ‘ diadem ’ from
Prakrit *makhula-, Skt. maknta, mukuta. Khotan Saka has a form
nearer to Skt., but with aspirate kh [= x) '■ murkhuta. A Saka form
*muhula or *mahula probably existed.
14. Agnean param, jxirn- translating Skt. pada ‘ position, rank ’
(adj. parno), Kuchean perne. Tock. Gram., p. 18, compared Sogd.
prn *farn.^ Khotan Saka phdrra < *farna- ‘ position ’ is used in
' Similarly r for Prakrit I <d = Skt. t regularly : kori- Skt. koti, Sogd. kwty,
Turk, kiddy, ku'lty, Khotan Saka kula ; Kuchean kakori, kdkoti, Khotan Saka kdkaula,
kdkold, Tib. kakola, Skt. kdkoti, kdkoli, Cf. also makara ‘ monkey * given by the
Kuchean saman m. w liei vgitm in the m ^ Fan-yu tsa-ming, ed.
Bagehi, p. 297, as Skt. [elsewhere marlcata]^ Khotan Saka makala ; Skt. I is replaced
by r in many other words {nira = ntla, vipuriya = vipulyCy ruka = loka) of this
vocabulary'.
* / is certain. Buddhist S<^dian does not always distinguish p and / (which is
possible by the alternative use of p and j3) but Manichean Sogdian has both p and /
and in this word gives /am, cf. Oss. /am ‘ luck
TTAUGABA
915
similar contents to those of Agnean and Kuchean, as Turkish uses
qut, in arrant quti ‘ position of an arhant ’ ; bur-xan quti ‘ position
of a Buddha ’ (the latter in both Buddhist and ilanichean texts) ;
sravaklar qutlari ‘ positions of Sravakas It is probable that Agnean
parn-, Kuchean perne are from an older Saka ^pliarna. Sogd. uses
farn in the same contexts.
15. Agnean mat^, Skt. makara ‘ sea-monster ’ (Toch. Gram.,
p. 62, note 1, Liiders, Zur Geschichte des ostasiat. Tierkreises, <SjBATT
1933, 1017, note 1). The attested Khotan Saka form is imqara E 25,
239 {klaisind magara ' the sea-monsters of klesas '), but a form
*malara- is possible and would then be the source of Agnean rndtcir.
Turkish wd’r, Mong. matar, Manchu. tmdari. Since in Khotan Saka
-t- in hiatus was probably not pronounced as -t-, it would be necessary
to imagine literary contact to explain the -t- in Agnean, after
the manner of French daseez from English da'.nsiy, das.iisig (dancing).
16. Agnean yamutsi- (92 b 2 kokaa .^paran ydnmtsm ‘ ruddy
geese, the spar-birds, ^ the parrots ’, 70 b 6 yumatstsissi kokdssi ‘ of
parrots and ruddy geese ') '' name of a bu'd explained by Poucha
(Tocharica VI, Arch. Or. 1933, 88 ff.) as the Chinese ^
'vng-miu-tsi ‘ parrot which is found also in a Sogdian text from
Tun-huang translated from Chinese {SCE 144, 315, 352) : >ymnctsy
*emutsi. Just as this proves contact with China for the Sogdian
translator so also for the Agnean language. According to Pelliot,
T'oung Pao, 1923, 317, and SCE. ii, p. 56, the use of ^ tsi as a suffix
is attested from the third century a.d. to the T'ang period in this word.
The texts in Agnean and Kuchean are largely religious works, but
in Kuchean other texts also, business documents and medical works,
are found, with which Agnean has nothing extant to compare. Hence
some of the foreign words appearing in Kuchean cannot be shown to
have existed in Agnean. The existence of a group of words from
Iranian in Kuchean and partly in Agnean is attested by tvdiikarai.
karkamassi, kimcit. aiikiras. arirdk, and with these, dmm and dsdnik
are probably rightly to be associated. The .s( of .^ostdnkdn, as indicated
above, does not exclude a loan from Indian or Iranian, but it may be
an indigenous Krorayina word. It can at least be said that the word
is known so far only in Central Asia. The words tmJiitr, nmhirsdii,
mdtdr. drsi and kdiak are equally Central Asian forms.
' Etymologically Engl, sparrow has been compared with Agnean spar- [' sparrow ’
is Skt. ca/(ita]. Would such a meaning suit here ?
VOL. VIII. PART 4.
59
916
H. W. BAILEY —
The importance for Dialect A is obvious. Here are words of Saka.
Central Asian Prakrit, Chinese and possibly Krorayina origin, whicL
could not be from Tokharistan. It is further confirmatory evidence
of the conclusion previously reached that Dialect A is truly a language
of Central Asia, the language of Agni.
[To this I am tempted to add the more uncertain evidence of the
word *okni deduced above, p. 899. The Sanskrit agni, the Saka
agnye, Chinese 'uo- and d (if this d is independent of the Sanskrit)
indicate a simple vowel and not a diphthong. Kuchean auk- is therefore
excluded. If *okrii has been rightly explained, it proves that the name
Agni comes from Dialect A, the language of Agni itself.]
CONCLUSIOX
It has been urged that toyam-toxara is the indigenous name of a
people of the dpoava-Snv’m region, who are later known in the
Bactrian region under the same name toxora. Their history thu>
coincides with that attributed by Chinese historians to the K
and the tuoxudid. A remnant of these is still known about a.d. 800
in the dpoava region. The native name toyam-toxara, and the name
dpoava-hrw”n indicate a polysyllabic language with voiced and
unvoiced fricatives. In Tokharistan they employed the Graeco-
Bactrian script. They had no attested ethnic or linguistic connection
with the northern cities of Agni, Kuci and Bharuka, from whose
language their own, as proved by these, the only certainly attested
words of the language — toyara, toxara, and dpoava — diverged widely
in phonemes. 1
A second people speaking dialects of one language extended
according to the linguistic evidence probably from Krorayina and
^ Reuter’s suggestion, Studia Orientalia {offered to K. TaUqvist)y 1925, 232-4, that
in Dialect A might represent the fricative Q would introduce an isolated fricative
into the language. Reuter himself recognized that no positive proof was to hand, but
he suggested three pieces of indirect evidence. (1) ts in aptsar- 8kt. apsaraSy and
samtsdra Skt. samsdra. Both these words, a fact probably not known to Reuter, occur
also in Khotan Saka with avdtsaray samtsdra (the latter also in KharogthI, BSOS
viii, 423, 427). The ts may be due to a Prakrit form with ts or direct from Khotan
Saka. Since Khotan Saka uses tk = $, it is definite proof against Reuter’s suggestion
of an interchange in Dialect A of ^ and 6 in these two words. (2) Interchange of
ts, tsts, tts, iss and ss. A value ts is equally comprehensible in these alternations. It is
also necessary to remember that ts > s may mark the later stage of the language.
(3) ts is treated as a single consonant and may be \vTitten doubled. Reuter assumed
that this excluded the value ts. There is an obvious error here in supposing that the
speakers of Dialect A thought of sounds according to modem phonetic analysis. The
case of c ~ ts indicates a different point of view. This c = ts was considered as a single
consonant and was written doubled in native words as bucc-assi ‘ what indeed ’ and
mdccek ipai ’ {Toch. Gram., p. 180, 192) and in Indian (Prakrit or Sanskrit) words.
TTATJGAKA
917
Niya in the south through Agni and Kuci to Bharuka in the north.
It still remains to discover if a single ethnic name existed for these
peoples, whether used by themselves or by foreigners. ^ Of their national
names two, Agneya and Kauceya, are known in Sanskrit.
West of Bharuka in the north and west of Niya in the south. Saka
dialects were spoken.
as in vicca-sim adj. to Skt. vidya [Toch. Gram.^ p. 54), as also cch in ucckist and
murcchdntu. That is, tts is written cc ~ tsts. The adoption oils for a sound felt to be
simple filled a gap in the Brahmi alphabet. Hence tsts means tts» as cc =- tsts means
tts, with which the alternative spelling tts agrees. The Chinese transcription of the
name of Kuci as kutsi shows that is was known there, and in Agnean ydmiitsi ts
represents Chinese ts. It must be noted also that in Khotan Saka kh th pJi are used
for fricatives (;^ B f), and that for Turkish x Hialect A also uses hkh {and hk) {Toch.
Sprachr. introd. xii, where x^tun should be read for qainn). Dialect A was therefore
aware of the convenience of this use of the Brahmi aspirates. To express B, th would
be expected according to system in Dialect A. We may note also the proposed
comparisons of the Niya Kharosthl Document kitsayitsa (a title, possibly ‘elder’)
with Kuchean ktsaitsahhe ‘ age ’ and of antkratsri with Kuchean akndtse, Agnean
dknats (= Skt. bdla ‘ignorant, young’), JRA^ 1935, 672-3. In these documents
ts is used for Skt. ts as in samvntsari. Here too we find th chosen to represent Iranian
6 in tkavamnaey ihavamnamae {BSOS vii (1934), 512), Khotan Saka thaiina * cloth
There is therefore no reason to conjecture that ts is B m Dialect A. The proof against
such a theory is positive.
* A word probably known also in Kuci since in the Kuchean Fan-yu tsa^ming
(ed. Bagchi, pp. 48 and 279, No. 537) tkacana ‘ cloth ’ is probably a misreading of
thavana. Here too th represents Iranian $.
^ It is hoped to take up this problem later. We have to recognize at least a cultural
connection between Krorayina and the northern cities. To this cultural unity belong
the three titles : (1) gauhira in a Sanskrit document from Kuci (Ltiders. Znr Geschirhte
und Geographie Ostturkestansy AS.4 U' 1922). corresponding to (fitiura of the
Kharosthl documents, (2) sothamgha in Agnean ^o-^tank-, (3) razfut in the Barcuq
Saka, corresponding to Kroravina cojhbo. Three naiiicH may induate even ethnic
connections. In the Sacu document edited by Konow we have in the region of
Tturpamni (Turpan, Turfan) the people ArgTiia in the phrase Arglhrd bi^d kamtha
‘ the town among the Argina With this may be com}>ared the name Argiya in the
Niya Kharosthl documents. A personal name in the Niya documents cimoln probably
appears in Khotan Saka, Ch. 00269, as the name of a people amuda associated with the
fivaiku.ra. It is possible to connect these with the cn?nul of Kasyarf, a people near
Bisbaiiq. It is admittedly impossible to prove they were not originally Turks, but it
is possible to see in them a people whom the Turks had absorbed. The third name is
ac^lh^. In the Niya Kharosthl documents occurs acuhiya amcutii oridnli as a personal
name. The same name is attested as the name of a king of Kuci. in C’hinese JliiJ ^ §2^
(i-t,nu~nzie quoted by Levi, Le “ Tokharien ", 22-3. Levi proposed to
equate *acuhi with Skt. arjunOy a phonetic equation which naturally seemed
doubtful to Pelliot, Tokh. 72, note 1. We may keep in mind also the still uncertain
Niya Kharosthl ogu and n ‘,wkn: of the Turkish colophon. There is also a possibility
that the Niya Kharosthl name kamjaka is connected with kancak, see above,
p. 913. In vocabulary’ we may. beside the reference to Burrow, JMAS 1935,
667 ff., note also Agnean ^tyok 'strophe’ lieside Niya Kharosthl sUyoka. .silyoga
' urkundliches t^chriftstiick ’ according to Luders, HSOS viii (1936), 6.54. It may
further be indicated that a section of the uo-sujit would probably solve the
problem.
918
H. W. BAILEY —
ADDENDA
I
The pages of the article “ Ttaugara ” were printed off in the early
part of the year. The study of the Central Asian documents ha.'
proceeded during the succeeding months, and it is possible to suggest
certain additions.
p. 883. The aksara ha followed by the two dots : indicates Turkish
y (and possibly q). Hence hvaihu.ra *hvaiyur ; -ttahi: , -ttaha.
-stahi, Turkish lay ‘ mountain ’ ; uha:, cf. Turkish oq ‘ division of a
tribe ' (not as p. 884, note 5, ‘ ehief ’).
bddumna may be compared with Turkish jnvdivn ‘ a people :
d for Turkish o would be normal, and it is possible to conjecture a
pronunciation *bodim (after a suggestion of Minorsky).
In 1. 78 it would perhaps be better to read dum va ucahi: spata.
supposing d(7»i to be the word dum in dum samgalaM of an unpublished
text, duni could be a place-name or tribal name (cf. perhaps the dvan.
a clan (?) name in the Sacu region, twice recorded in a Tibetan
document, JRAS., 1927, 827); ticahi: could be a Turkish *ucay
(? *ucaq) ; and spata could be careless writing for spata ‘ general',
which is attested both before and after names.
sfdya could equally well mean ‘ Sogdians ’ and so be added to the
list of forms in BSOS., vi, 948. A Sogdian colony is known at Lob-nor,
Pelliot, JA., 1916, i, 111 ff.
Y. ilinorsky has suggested a comparison of imju with Turkish
incii ■ appanage ' ; of adapahutti with Turkish alpayut ; of
tturhi bayarkdva with liirk and bayirqu, a tribe of the Uighurs ; and
of hdtti bara with PJ a tribe of the Tblis federation.
p. 886, line 9. Read -8 for -6.
p. 890. Read pr'tyk'pwd for prdy’pwd.
p. 895. Read Nagardhdra.
p. 896, note 3. The kingdom of Kuci {Kuci raja-) is known in the
Niya KharosthI documents.
p. 897, note 3. A. Freiman has written to inform me that document
no. 15 of the Sogdiiskii sbornik is written in Turkish runic script, but
in a yet imdetermined language.
p. 899. Read also Aaot.
p. 902. It may be useful to note that the “ Four Great Divine
Kings ” are the four Lokapalas, Vaisravaua, Dhrtarastra, Virudhaka,
TTAUGARA
919
and Yirupaksa, famous in the seventh chapter of the Suvarna-
bhasottamasutra.
p. 906. The Skt. drya appears in the Sanskrit-Tibetan formulary
(ed. Hackin, Formulaire sanscrit-tihetain, 1924) in the forms a-rja
( j = dz) and a-rjya {j = dz), and tiryak in the form tri-ja-ka. Hackin
refers also to an unpublished Uighur Turkish text with ar'ja, p. 102.
p. 907. Read Maitreya-samiti-ndtaka.
p. 913, note 1. kancak is called ga-hjag in Tibetan documents,
F. ^Y. Thomas, Tibetan Literary Texts and Documents concerning
Chinese Turkestan, pp. 118 and 133. Kasyari hi, 181, 8, quotes s
‘ drinking, carouse by night ’ as a Kancaki word, see Brockelmann.
Hsia Major, h, 1, 121.
p. 914. Sogd. pt^y ’Scy’H’k ‘ worthy of honour ' (’scy- derivative
of arg-) renders ‘ arhant ', Dhyana Text 14, Benveniste, JA., 1933,
ii, 215.
p. 916. It would be safer to say ; " loyara and dpoava do not
exclude a polysyllabic language.'’
p. 917 (in note 1 to p. 916). Both ts and tsts represent Chinese
ts in ydmutsi. tsts for ts has also crept into Skt. texts from the same
region, see JRAS., 1912, 355, and 370 : utstsuke.pi tv anutsukd{h).
p. 917, note 1. Read ’ the town among the Argina ’, not ’ towns ' ;
kamtha is singular.
II
The present opportunity should be taken to supplement two other
articles in this volume.
(1) It was noted too late in the “ Fragment of the Uttaratantra
in Sanskrit ”, p. 86, that ukdtta, 1. 22, had been misread for uskdtta,
which in Saka means ‘ above '. It is therefore clearly a direction to the
reader to transpose the two verses of Sanskrit, as Dr. Johnston had
already seen to be necessary, p. 87, note 11. For uskdtta, cf. Siddhasara,
5, V. 4 : amstarn astam uskdtta uskdtta, Tibetan phyi-ma phyi-ma la
sna-ma sna-ma.
(2) Iranian Studies Y.
Further work at the Khotan Saka texts has brought some clearer
views.
agane was wrongly explained as ‘ powerless ’. It proves to corre-
spond to Skt. uru- ' thigh ', vaksas- ‘ breasts', Tibetan ro» (1 for 6ra»
‘ breast ’).
920
H. W. BAILEY —
ayala is Skt. ayogula in the story of Mahaprahhasa in the
Kalpanamanditika, fol. 167, ed. Liiders.
dlde is (in a medical text) ‘ languor
eysd {-d is frequently for -am) < *alzana- NPers. arzan (after a
suggestion of Morgenstierne).
haugd, E. 2, 49 ; 17, 12, hulgo ‘ soft h
kaba, Chinese ^ kdp ‘ decilitre ’, Turkish qav, F. W. K. Muller,
Uigurica, ii, 82, cf. Kuchean kaui.
nuska, anuta < ni-yauk- (according to a suggestion of
Morgenstierne) or ni-auk- : -sk- unvoiced group indicates k not y.
pdja, E. 4, 26 ; 12, 35. patargya.
phamnai, Skt. tdlu ‘ palate ?< older Saka *p}iatana-.
simga, Chinese Th ^^>^9 ‘ Ufre ’, Turkish sing, Muller, Uigurica, ii.
82. Both dm simga ' two simga ’ and dva sirnga halai ‘ two and a
half simga ’ translate Skt. prasikatn. Since we have also dvi tndcdmgyi
hold ‘ two macamga and a half ’ for Skt. prastham we may infer
that 1 sitnga = 1 mdcdmga. Then mdcdmga will be the Saka word
corresponding to the foreign shnga : mac- < *matc- < *mdtak-, cf
Saka dee ‘ water-bird ’ < *dtc- < *dlac- or *dtic-.
ysuma ‘ broth ’, cf. Pasto zwamna ‘ soup, broth ’, Morgenstierne
EVP., p. 104 (? ica < u < au) to zau- : zu- ‘ pour out
The following errata have been noted : —
s.v. gvir- : gudd 3 sing. fern, (not 2 sing.),
s.v. gyasta- : read ysdndhdiiu.
Read hamphlsdnd.
Read iharstd with t.
s.v. khdje : read E 6, 90.
s.v. kutqnd : read Skt. kuU-.
Read naspastqme.
Read pdraingd.
s.v. pdrysa : read E 2, 50 parysa.
s.v. pilhi : read puhyai.
Read sambajdtu.
s.v. ttumgare : read tvdhkaro, cf. the plate JA., 1911, ii, facing
p. 120 : Ivdhkarai with v.
tv{ine, produces flesh. Skt. brmhanu-
Read ustamjdna.
s.v. t>m : read enema.
TTAUGARA
921
POSTSCRIPTUM.
This may be the place to refer to two further contributions to the
solution of the vexed problem of Central Asia. G. Haloun read a
paper (which will, it is hoped, be published soon) to the German
Oriental Society in Bonn in September this year. Pelliot in T'oung
Pao, 1936, has contributed further matter of importance in an article
entitled A propos du “ toMarien From his discussion of the two
Chinese characters ^ and found in transcriptions of the name of
Agni, it is clearly safer to leave them aside at present. There remain
Bareuq Saka aghye (whence a nom. sing, agiii- could be deduced)
and Chinese transcriptions indicating The initial vowel of the
native name is uncertain ; one can conjecture a, o or e. If we give
greater credence to the more flexible Brahml script,' we may suppose
a word ending in -gni or in the system of Dialect A -Hi. Should it
be desirable, a connection with Skt. a/ii-, or the form with nasal as in
Lith. ang'is ‘ snake ’, may be maintained.
It may be noted, in reference to p. 265 touching Agni and Agnean,
that “ Agnean ’’ is simply an Englishing of Skt. agneya which the
Agneans themselves used, just as " Kuchean " is an Englishing of
the Skt. Kauceya, Kaucya used by the people of Kuci.''
The name of Tun-huang (discussed here, p. 262 if.) will need yet
further consideration. In this connection Karlgren's study in the
Ts‘ai Yiian P‘ei Anniversary Volume, 1933, <So/«e Turkish transcrip-
Uons in the light of irregular aspirates in Mandarin, will be of service.
It \Ndll not, however, be useful to start from the Chinese transcription.
Sogdian Brw"n interpreted by Greek Opoava can alone be accepted
as authoritative. In Sogdian words Sr- may represent Sr- (or Ir-), Old
Iran, dr-, or dr-, Old Iran. dr-. The more precise Greek script (which
was not unknown in Central Asia) could define the Sr- here as dr-.
I am reminded by Hansen of the tuyr'kcny in the Kara-balgasun
inscription, § 19, and by Minovi of the tirh'ryk of the Dra;:^t I asorik,
§42.
' The name ta^y:ic in Turki.sh. ^ Vnk-h‘udl (or ^ in Chinese,
(Ueek TouyaoT, similarly offers fiy and kh'. \\*e should prohaitly prefer to trust the
form in Sogdian script.
- Three unimportant points may he indicated here. On p. 261 the reference to
Switzerland is due to a misunderstanding of the phrase "identical language ’ ;
Iswiss kit ■ been ' is sufficiently different from German On p. 264 the earlier
explanation of Syriac tkivnln is repeated without reference to the essential difficulty,
the absence of ic in the first syllable. On the same page, lines 3-4 are not strictly
correct, since neither of the two erroneous forms (due to an error in copying, since
no originality was sought in the Chinese conjectural reconstructions) entered into
the following discussion, and t uo^udld was attested by the third Chinese transcription.
Hvatanica
By H. W. Bailey
TT has seemed desirable to make known at once certain information
of interest to Central Asian studies contained in Khotan texts of
the British Museum and India Office. This information may here be
conveniently grouped under the heads (1) the animal cycle of twelve
years, (2) the names of the months and seasons in Khotan, (3) dates and
royal names.
Since the forms quoted in the following pages are from various tj'pes
of Khotan documents, it will be well to prefix the statement that three
linguistic stages can be easily detected in the extant Khotan
manuscripts. The first stage is represented by texts such as those
pubhshed by Ernst Leumann in Das nordariscke [saifsc/ie] Lehrgedicht
des Buddhismus, a stage which itself was preceded by an earher period
in which the orthography was fixed. The second stage may be found in
the text of the Vajracchedika, edited by Konow in Hoernle, Manuscript
Remains of Buddhist Literature from Chinese Turkestan, in which the
inflections are already much- reduced, and the orthographic system, still
good, is much modified. The third stage is met with in many rehgious
and other texts, showing confused knowledge of the old inflection and
much interchange of the vowels. A good specimen can be seen in the
Itinerary edited in Acta Orientalia, xiv, 258-267, which is of the tenth
century a.d.^ I am indebted to Professor Sten Konow for some
valuable suggestions.
* I take this opportunity to give an explanation of the word gara- in lines '2S and
33 of that text. In the manuscript of Khotan, written by the Pandita Mo-rgu-bde-sil
and translated in F. W. Thomas, Tibetan Literary Textn and Documents concerning
Chinese Turkestan, 1935, p. 321, we read : “ In the Li language ' hut ’ is hgo-ban.”
Li is Khotan. The word hut ” in the Tibetan text is, as Professor Thoma.s has kindly
informed me, spyil-bu, that is, the Sanskrit trna-kulira, a hut of grass or straw. Mow
hgo-han could represent gaca- of the Khotan Saka text. In the Itinerary we have 28 :
sagija yai’am jsa, and in 33 ; jardjsa " with gava- of stone It is probably safe
to conjecture that the author, familiar with thatched huts, felt the need to insist
that here the huts or cells (gava-) were of stone. On another word, spa, beside the well-
known 'a-ma-ca, in this Tibetan manuscript, see below, p. 934. It may also be of use to
suggest that the ka-rots of the name hjah-mo-ka ka-ron (F. W. Thomas, loc. cit., p. 117),
the first foundation of the Sarvastivada school in Khotan, is the word karana-, nom.
sg. kardtn of the Itinerary, ed. Acta Orient., xiv, line 12. In line 17 tharkye is probably
represented by tharka in the phrase tkarka mijsa, Ch. 00265, 37, “ marrow or kernel
of the tharka ” in a list of plants, following iramde " castor-plant ". It is not yet
identified. A passage in which bada- is parallel to jinave (Skt. janapada) has been
noted. Two errata on page 266 of the same Itinerary should be corrected : read
janub and yamtadd.
924
H. W. BAILEY — ■
I
Brit. Mus. Or. 11252 (1) presented by Sherriff, late Consul oi
Kashghar, probably from the region of Khotan. It is a roll of coarse
brown paper, 152 x 28 cm., complete in length, but somewhat broken
on the left side (obverse). Obverse 56 lines of cursive Khotan Brahmi
script. Reverse 1 line. [Itabcs in the transcribed text indicate lost
syllables.]
Reverse
12 si salya baya u garane Ttcelve Year Leaders atid their
Influences.
Obverse
1. salya haya 12 cu siri diri f
varai bva dvasi salya baya sau
2. hada hdylda khu sau hada
basti yanidi ttl sa sa sail
bayidi
3. . . pasa mula
4. sva ssava guhi
5. . . brrunasti muyi
6. ... urmaysda sahaici
7. . . mase na ysai ste
8. . . stena mase saysdi
9. . . da mase asi
10. , paryai sva hada pasi
11. . ma ha ra da badi makala
12. . . na-yasari krregi
13. . *rdunari hime tta sve
14. . ma hveram pasi
15. . . tt(i) salya baya sau kala
bayidi buna buna
The year leaders are twelve. Know
therein what good and evil ivill he.
The twelve year leaders lead
one day. As they can lead one
day, so they lead each year.
[10.0 p.m.J . . evening . Rat.
[12.0 p.w.] midnight . Ox.
[2.0 a.w.] . . . dauming . Tiger.
[4.0 a.m .] .... sun . Hare.
[6.0 O.W.] Dragon . ■ ■
[8.0 a.m.'] Snake.
[10.0 a.m.] .... Horse.
[12.0 noon] . . midday. Sheep.
[2.0 p.rn.] time. Monkey.
\fl.0 p.m.] . . . Cock.
[6.0 p.m.] . . . Dog.
[8.0 p.m.] . . . eating . Hog.
. . These Year Leaders lead one
period of time ....
Reverse
bdyd, pi. to *bdyai ' leader ', cf. bdrai ‘ rider '. bdy- ‘ to lead as
Skt. nayati, of control by supernatural agents.
garane, Skt. karana, cf. ggaranisdstra-, Skt. karanasdstra.
Obverse
7. y.sai ste unknown.
12. -vamri, cf. pamra- ‘ evening
HVATANICA
925
16. II padauysa mxila mula salya First the Rat year. A man is horn,
hve ysaiyi muysga-jsml liime he becomes short-lived and
u Saudi . . .
17. . . bariji ni siri [hijhimare u ... crops are not good and the
live yaulajsi hime u sahautti man becomes evil (?) and is not
happy.
18. ni hime cu jvidi ba-jsinya When they live they are short-
himare u travi pbaraka himare lived and abound and
u hlni pha- armies are many and evil.
19. raka wiasiri
20. ... sva ssava cu guhi salya . . . Midnight, when it is the Ox
live ysaiyi sahautti hime u year. A man is born, he becomes
cuai [ma] pu- happy, and if to him sons
21 . ra himade bisi sahautti are born, all are happy and as to
himare u cu barlja-kerai sirai a soicer of crops, he is good
hime u stu- for him, and cattle
22 . ra phaiaka, himare u girye abound and he can do good buying
para siri yudi yarnde u uci and selling, and for him there is
jsai p^pi u dai fear of water and fire.
23. || muyi salya hve ysaiyi year. A man is born, he be-
bihaysi hime u jauysi u barijai comes a traveller (?) and warrior
ni siri hi- crops are not good for him,
24. me u haphara-sala hime u and he talks folly and is long-
i^ysa-jsini lived.
25. j| sahaici salya hve ysaiyi Hare year. A man is born, in all
bisina sahautti hime muryau he is happy. W ith money, cattle,
jsa sturyau jsa asau jsa horses.
16. haudi, perhaps cf. hada- ‘ small .
17. 6 ar 7 ji possibly an adj. fern. -Twja from 6ara-* crops , used as
a noun.
21. barlja-kerai, cf. Or. 11344 ( 8 ) 8 : hvaindi 24 middm jasti hlya
kdinjsata-kerd u ganama-kerain ‘ 24 men, His gracious Majesty s
sesame-sowers and wheat-sowers .
22 . girye para, to gir- ‘ to buy and para- to sell .
23. hihdysi, see Acta Orient., xiv, 263.
24. haphara. E 13, 89, samgganye yande bodhisattva haphdre.
haphdra- (ha- < *fra-) beside dphdra- ' disturbance .
926
H. W. BAILKY
26. . . bisau jsa bisina suhye
hime u baysai pamdi ni himye
u bedai asye
27. himdT\ u kbu barija-keri
sirai ni bime u bisana vasu
bime aysmuna
28. rasu bime grrabina mide
29. ;| na salya bari pba hime u
bramtbi tsidi u utca pba bime
u bini
30. pba tsidi bada bamyari u bisi
pvanidi u dai pitti daina pyani.
31. .CM bve ysaiyi buysa-jsini
hime
32. j[ saysdi salya bvai ysaiyi
utcai pba hime u bari pba hime
u ba-
33. rljai siri hire ni himare u
acbai pba hime bvamdi mi-
ramre u
34. daina pvarnne cu bve ysyane
buysa-jsini hime
. . ., houses, in all he is happy,
and for him there is no distant
travelling and on him itching
diseases {?) come, and as to
a scnver of crops, he is not good
for him and with his tongue he
is evil ; in his mind he
becomes evil. He dies through the
influence of a planet.
Dragon year. Much rain falls
and winds blow and ivater
abounds and many armies go out.
times change and all are afraid,
and fire falls, there is fear of fire.
When a man is born, he is long-
lived.
Snahe year. A man is born. For
him water abounds and much
rain falls and
crops are not good things, and
disease abounds, men die and
there is fear of fire. When a man
begets children, they are long-
lived.
35. ]j asi salya bve ysaiyi baysai
anai hime u kbu biri masa yidi
36. kiri ni parsti idi u hvatnda
sani pba bimari u ba du haysidi.
37. vasuni acbai bime mirare.
Horse year. A man is born. For
him there is distant travel (?) and
ivhen he makes a thing . . .,
his % 1 'orks are not . . ., and
human enemies abound and they
work ....
Evil disease arises, they die.
26. dsye, gsyeni in 40, cf. dsyam, Skt. kandu ‘ itch ’.
28. grraha-, Skt. graha- ‘ planet ’.
HVATANICA
927
38. 11 pasi salya hve ysaiyi
sahautti hime punaudi bise
hirina sa-
39. hautti hime jsarina u muryau
jsa u achinudi hime u nmysga-
jsi-
40. ni bedai vasuna acha himare
u asyeni sarbidi u vrrani ka
41. . . sti kuri mijsl inirare u cii
pura ysanide muysga-islnya
hi made
Sheep year. A nwn is born, he is
happy and virtuous. In every
affair he is happy,
•with corn and with money, and
he becomes diseased and short-
lived.
Upon him, come evil diseases, and
itching diseases (?) arise and
ivounds for him . .
. . . perverse wives die for him and
■when they bear sons they are
short-lived.
42. ;i makala salya hve ysaiyi
haysai samda anai hime u pfirai
pha himare
43. bisi hirina siri hime bisai pha
himari u asa u khu hiri masa
44. . . -i *fiai arhl idi bisi
jsimdi u puryau jsa aspata ni
bye-
45. hldi u tsidi a mirare
Monkey year. A man is born. For
him there is travel (?) to a far
land and for him sons are many.
In every thing it is good. For him
houses {or servants) abound and
horses and when he makes (?) a
thing . . .
. . . for him they are restricted.
All are slain and in sons they
attain no refuge,
and they go away or die.
46. ! krregi salya hve ysaiyi cuai Cock year. A man is born. When
pura himade bisi muysga-jsinya for him there are sons, all are
hi' short-lived
47. made u sahautta himare and are happy. With barley,
rrusa ganani asa muri pamne ivheat, horses, money, with every-
hirina sahau- thing they are happy.
39. dchinuda- ‘diseased’, and in 54, cf. plruda- ‘with worms’,
padajudi ‘*having rank’ {padamja-). aysdada- ‘watched’ < Old Iran.
*azdd-krta- (wrongly explained BSOS., vi, 77), beside dysda-yandka
‘watcher’. Ch. c. 001, 998; aysdagarai ‘watching’, dysdarrja sb.
‘ watching ’ ; pajsamadi ‘ honoured ’ (corresponding to Skt. pujate)
< *pati-)ama-krta-.
44. drhi, possibly a later form of ahrri < *d-draxta-, Skt.
stambha-, Tibetan dban-med-pa ‘ powerless ’.
928
H. W. BAILEY —
48. tta himari u kirl ni parstl
yanidi daina pyani u ^larbi
hlsl-
49. di jsari hvaridi
50. II sva salya hve ysaijd
muysga-jsmi hime u dikhau u
cuai mu pura
51. ysanlde bisi muysga-jsmya
himare u bi^ yina u haphara-
sa-
52. Id ^t nauqla u salarbi hisidi
jsari hvamridi
53. ji cu pasi salya hve ysaiyi
muysga-jsmI hime u purai bisi
muysga-
54. [ . ] jsifiya himari u
achinucja ksira bisi achai hime
phara-
55. ka u hlni ni tsidi u bisana
satta va^na himari u salarbi
56. hisidi jsari hvarnridi
arid for him they do not niah-
and . . . his u'ork. There is fear
of fire and the locusts come and
eat the corn.
Dog year. A man is born, he is
short-lived and unhappy, and ij
they bear sons,
all are short-lived, and . . . .and
speak folly and are
. . . and the locusts come and eat
the corn.
Hog year. A man is born, he is
short-lived, and for him all sons
are short-lived and diseased.
In the whole country is much
disease,
and armies do not go away and
with the tongue people become
evil and the locusts come and eat
the corn.
Or. 11252 (1) accordingly supplies the complete list of the twelve
names of the animal cycle, some of the names being repeated. The
names compared with other known Central Asian lists may be tabulated
as follows : —
Khotan iSaka
Sogdian
j Krorayina Prakrit
Sanskrit
Kuchean
miila, mula, mula
, mws
1 mu$ka
mantilya
arsakarsa
guhi
y
gava
govr?a
okso
muyi
myw
1 vyagra
yyaghra
mewiyo
sahaici
' Xry'vsy i
sa^ka
sasa
?ase
na
n'k I
nag'a
naga
, nak
saysdi
kyrmy ]
jamdunanica
jantunah
auk
asi
1
; as])a
asva
yakwe
pasi
psyy 1
pasu
pa4u
saiyye
makala
mkr’ i
i makad'a
markata
mokomske
krregi
mryyy
1 kukud'a
kukkuta
kraiiko
sve, sva
kwtv 1
, bvana
svana
ku
p^si, pasi
k’s '
! sug'ara
sukara
suwo
48, 52, 55, salarbi ‘locusts ’ cf. Ch. 1, 00215, b 53 hina jsa pvena
harlysa darabaiksa .... salarba . . . achd ‘ fear of armies, trembhng,
famine . . . locusts . . . diseases.’ E 2, 66 -sscdarba.
HVATANICA
929
To the names should be remarked : —
1. ‘ rat mu- < *muz- Old Iran. *mus (Skt. mus-, NPers.
mus), with suffix -la.
2. guha- translates Skt. go-, Tib. ba-lan ‘ ox '.
3. muya- ‘ tiger ’ for *nmya- < *rnauya-. Sogdian myiv ‘ tiger ’
can be read *mdy- or *muy- (earlier transcriptions *meiv, *mTiv
assume that the Sogdian orthographic system was identical with the
West Iranian. It is well to remember such Sogdian forms as
sywhyk ‘ Sogdian ’ and rywsn ‘ bright West Iranian ricin).
Kuchean has mewiyo (Liiders, Zur Geschichte des ostasiatischen
Tierhreises, p. 24), mancyai ‘ tiger ’ (Levi, Le “ Tokharien ”, p. 23),
mewya ‘ tigress ’ (Liiders, loc. cit., p. 24), which agree in form with
Khotan Saka and Sogdian. There would be no need to maintain the
improbable connection with Chinese ^ miiiu, mau ‘ cat ’.
4. sahaici ‘ hare attested in the medical text Siddhasara in the
list of edible flesh, 16, v. 5, sahaicd hiya gusto, Tib. ri-bon gi sa ‘ flesh
of hare ’.
5. na, Skt. naga- in Khotan Saka also naga-, ndta-.
6. saysda- ‘ snake ’, in Leumann's texts ssaysde. The Krorayina
Prakrit jamdummca and Skt. jantunah may perhaps include the snake
with the ‘ insect, worm ', as does Mid. Pers.
Xrafstra-. We may note also an example of this inclusion of snake with
worm in Ossetic kalm ‘ snake, worm NPers. kirm ‘ worm ’, Sogd.
kirrn- ‘ snake The lists would then all be in agreement for this year
of the cycle.
7. asi, older assa- ‘ horse ’.
8. pasi probably applies to either ‘ sheep ’ or ‘ goat ’. Av. ■pasu-
refers to small cattle in contrast to staora- ‘ big cattle '. The special
names are given in the Siddhasara 16, v. 3, mina gusto ‘ sheep's flesh ',
Old Iran. *maisa-, and buyslna gusto ‘ goat’s flesh ’, Old Iran. *buza-.
In Ch. 0048, 1, pasa salya (the pa is broken but certain) was misread
ssa salya by Hoernle, JR AS., 1911, 470, and so caused trouble to
Konow, Acta Orient., vii, 66, and Liiders, loc. cit., 26, no. 3.
9. makala ‘ monkey ’, corresponding to Skt. makara of the
Kuchean Fan-yu Tsa-ming, ed. Bagchi, no. 905, p. 297. It was wrongly
read madala (?) by Hoernle, JRAS., 1911, 470, misleading Konow,
Acta Orient., vii, 71, and Liiders, loc. cit., 27. In the cursive Brahmi
script of the Khotan documents da and ka approximate graphically,
but never so closely as to be confused.
10. krregi ‘ cock ’, frequent, older krrimga-.
930
H. W. BAILEY —
11. sve, h'd ‘ dog elsewhere svana, nom. pi.
12. pfci, pasi ‘ hog ’. In the Siddhasara, 16, v. 4, pasa gu-i
renders Skt. mamsam varaham ‘ boar’s flesh ’. Morgenstierne li<
recognized in it Old Iran. *parsa-, Latin porcus, corresponding to Kuri
purs (see Acta Orient., i, 277).
It is unfortunate that the names of the two-hour divisions of th
day are so badly broken initially.
II
The second text is an excerpt from the Khotan translation, Cii. ii.
002, of the Siddhasara of Ravigupta (sixty -five folios have survived).
In Indian medical texts, the physician is instructed to observe tin
seasons and the months in his diagnoses. To this we owe the followint:
complete list of the names of the Khotan months and seasons.
Siddhasara 3 r 4 — 4 r 1. Plate VI.
ttye hlya piskista jsa tta biraysare *. rve ksa biraste ||
hamdyaji masta u raruya si hamamna rva ste '
cu ttamjara u brakhaysdya sa pasamjsya rva ste
cu mutcaci masca u mumnamja si ysumamna rva ste ||
cu skarhvara masta u rrahaja si ysumi pastyi bisa rva ste |1
cu cvataji masta u kaja si pasalya ||
cu hamariji masta u simjsiinja si nastyi pasala bisa iva ste 1|
They are so explained according to its divisions. Six seasons are
distinguished. The month Hatndyaji and Raruya are the summer
season. Ttamjara and Brakhay.3dya are the autumn season. The month
Mutcaci and Murnhamja are the pointer season. The month Skarhvara
and Rrahaja are the season at the end of winter. The month Cvataji and
Kaja are the spring season. The mmith Hamariji and Sitnjsirtija are
the season at the end of sjrring.
Skt. Text.
vacat
pravrd nabho-nabhasyau ca
isorjau ca saran matau
Tibetan Translation
dehi rgyun gyi rnam-pa dbye-ba
ni I dus dhigs mam drug-tu
bsad-de |
de-la dbyar zla-tha chun dan ston
brla ra-ba gnis ni dbyar g}d
dus-so I
ston zla hbrin-po dan ston zla-
tha chuD gnis ni ston gyi
dus-so I
HVATANICA
931
Skt. Text.
marga-pausau ca hemantali
sisirau magha-phalgunau
vasantas caitra-vaisakhau
nidaghah suci-sukra-bhak
Tibetan Translation.
dgun zla ra-ba da» zla hbrio-po
gnis ni dgun gyi dus-so |
dgun zla-tha cbun dan dpyid zla-
ra-ba gnis ni dgun smad kyi
dus-so I
dpyid zla hbrin-po dan dpyid
zla-tha cbun gnis ni dpyid
kyi dus-so |
dbyar zla ra-ba dan dbyar zla
hbrin-po gnis ni sos-kahi dus-
so 11
There follows a section absent from both Sanskrit and Tibetan
texts : —
patch sau pacada ^
cu hamdyaji myarn masti amna auda ttamjeri myam masti hime
si hamamna rva
ttainjeri myam masti arnna auda mutcaci myam masti bure
si pasarnjsya rva
mut^ici myarn masti arnna oda skarhveri myarn masti bure
si ysumaiiina rva
skarhveri myarn masti arnna aiida cvataji myarn masti bure
si rrastya ysumarn bisa rva
cvavaji myarn masti arnna oda hamarlji myarn masti bure
si pasalya rva
hamarrji myarn masti arnna auda harndyaji myarn masti bure
si nascyi pasalarnjsya rva
Afterwards a second division : —
From mid Hamdyaja to mid Ttamjara is summer.
From mid Ttamjara to mid Mutcaci is autumn.
From mid Mutcaci to mid Skarhvdra is winter.
From mid Skarhvara to mid Cvataji is the end of winter.
From mid Cvataji to mid Hamarlji is spring.
From mid Hamarlji to mid Harndyaji is the end of spring.
This second account is perhaps independent of the Indian tradition.
We are, however, hardly justified by this alone in conjecturing that the
month Harndyaji was the first month of the Khotan year.
Two short following paragraphs should be noted here for the names
of seasons.
VOL. VIII. PART 4.
60
932
H. W. BAILEY
khu umiaysdi hauda rricliam pa
jsate u rravye pa jsate dvyam
pamda vam pastame j sa varsa-
vaysa ^ u ysumamna u
kamam
Skt. Text.
ete ca varsa-sltosna ra\’i-vartma-
dvayasrayah
When the sun goes northwards
and goes southwards, because of
his moving on two paths, then
are the rains, and winter and
summer.
Tibetan Translation.
ni-ma byau phyogs-su hgro-ba
dau I Iho phyogs hgTO-zio lam
gSis-su hzug-pahi phyir de-la
char hbab-pa grau-bahi dus
daD I cha-bahi dus-su gyur-pa-
ho II
varsavaysa u ysumam u hamam
ttavamdya •v u ^ilisam u beta
Vi bure pasa u pasala u hamam
aphirare
Skt. Text.
cayo varsa-himosnesu pitta-
slesma-nabhasvatam
kopah sarad - vasantambuvaha -
kalesu klrtitah
The rains and winter and summer ,
the bile and phlegm and wind
gather {?), autumn and spring
and summer they are agitated.
Tibetan Translation.
char hbab-pa dao | grao-ba dan |
cha-bahi dus-su mkhris-pa
da» I bad-kan dau rlu» rnams
gsog-cin I ston da» dpyid da»
dbyar hkhrug-par gyur-te |
This text has therefore preserved for us all twelve month names m
regular order and in good orthography. The names are also known in
documents in various later spellings. It will hardly be necessary to
repeat here earlier misreadings. They can be found in Hoernle’s article,
TEAS., 1911, 471.
The months with Sanskrit equivalents are therefore as follows : —
1. hamdyaii sravana I,
„ . , , - T 1 -hamamna rva
2. raruya bhadrapada |
3. ttamiara asvina |
4. brakhaysdya kartika
5. mut^ci marga&sa )
~ . ’ rysmnamna rva
o. mumnamja pausa j
^ Pali vassdvasa-, Niya Xharo§thI doc. varsavasa-.
HVATANICA
933
7.
skarhvara
magha
8.
rrahaja
phalguna
9.
cvataji
caitra
10.
kaja
vaisakha
11.
hamariji
asadha
12.
simjslmja
jyestha
Variant forms : —
j nastya ysumam bisa rva
! pasalya < rva >
pasalanijsya rva
ttamjeri (gen. sing.), ttaujimri
skarihvari
cvavaji cuvija [«of cuvijsa, Acta Or., vii, 66J
Khotan seems to have known four seasons : pasald ‘ spring
hamdm ‘ summer pasa “ autumn ysumam ‘ winter '. To translate
the Indian text two additional phrases were invented to correspond to
the Indian division into six seasons.
Afragmentary calendar is also given by Leumann, loc. cit., ii, p. 357,
preserved on fob 2906 of MS. E. It can now with these new data be
more successfully interpreted. For completeness it may be added here.
sya brrarnkhaysji masti stamna audi rrahaja inasti buri
hemmamntai samai tcahau masti
4ya rrahaja masti grismi samai ....
buri varsi samai sa masti
tti va sya harndyaja masti stanma audi sy^ brrarnkhaysji
masta buri dirgha-varsi samai drai masti
Here, too, the reckoning is from mid-month (sm = middle). But
the seasons differ greatly from the other two accounts.
Ill
Dates from the Khotan Saka documents have already been
discussed by Hoemle, JRAS., 1911, 469 ff., and Konow, Acta Orient.,
vii, 66 ff. Other dates which have since come to light may suitably
be given here.
It is particularly useful to have both a name according to the cyclical
animal years and the serial number of the year. Of such double datings
Konow had only two. Five new cases can now be added.
^ »Tataka-stava 5 r 3 pasarnjsye purre hivi tti^ ‘ splendour of the autumn moon
also contains the adj. form to pa^~.
934
H. W. BAILEY
; First year
I of k§una
Ch. i, 00216, a 43 : puhye ksuamni pasa silya 'fifth
ksuna, Sheep year
Ch. 00269 : guha salya hau ksani ‘ Ox year, seventh ksuna \
S 2469 : 24 ksumni asi salya ‘ 24f6 ksuna. Horse year
Sacu Document (ed. Konow) : 14 ksumni krrimgi salya
‘ 14^A ksutia, Cock year
Ch. 00275 (Hoernle, Manuscript Remains, plate v, fol. 1) :
30 myi ksauna guha salya ‘ 30th ksuna. Ox year ’.
Ch. c. 002 : naumye ksauna asa salya ‘ ninth ksuna. Horse
year ’
Ch. 00272 : tcaulasamyi ksaunapv aisa * salya ‘fourteenth
ksuna, Hog year
' Hare
Sheep
Sheep
I Monkey
I Monkey
[
Hog
Dog
The word kmna, which is found also in the Niya Kharosthi
documents, as in no. 715, amgvaka devaputrasa ksumnamnii ‘ in the
ksuna of Amgvaka the devaputra and in Kuchean kmtn (Levi.
“ Tokharien ” B, langue de Koutcha, pp. 8 IF., and Le “ Tokharien ’ ,
p. 23) has already been explained as “ regnal year ”. The new dates
with cyclical names give five starting-points for ksuya in Khotan
Saka documents. This fact seems to exclude the conjecture (Konow,
Acta Orient., vii, 71, 76) that the ksuna were contemporary eras of
different regions. We have evidently to do with regnal periods
reckoned from each new king’s accession. The kings are certainly
in some cases kings of Khotan, and probably all the ksuna must be
referred to them.
Since, then, the kmna are an indication of the length of the reigns
of the kings, other documents giving serial numbers of the years
without cyclical names become important.
A series of documents exists which can be shown to have come
from the Khotan region, largely concerned with the official business
of the General Sudarrjarn. That is in Khotan Saka spdta suddrrjdm,
and spdta < *spddapati-, beside which the later form spd is found,
corresponds in the Sumukha Dhdrani to Tibetan sde-dpon. In one of
these. Or. 11252 (37) B, a document of Samdara addressed to the
General Sudarrjarn, the name and title are repeated at the foot of the
document in Tibetan script as spa : sor : zon : la ‘ to the Spa Sor-zon ’.
^ piaisa in a badly written text for pdsa.
HVATANICA
935
We have here sor-zon = suddrrjam and spa = spa (spat a) ‘ general
The 0 of the name resembles the frequent examples elsewhere, as in the
Tib. bohan, Khotan Saka vdham of the royal name T im-vdham, or
in the Tibetan text translated by F. W. Thomas, loc. cit., 303-323.
One may note especially p. 321, ’or-non, the name of a Sangharama,
in which will be recognized Khotan Saka *iirijam, older itryana-
‘ garden ’. The o is incidentally of interest for Khotan pronunciation.’^
It may be useful to indicate that a man of Khotan called Sar-ion
appears in JRAS., 1930, 51-2. W'e shall probably also recognize in this
spa ‘ general ’ the word spa, evidently a title, of the same Tibetan text,
p. 320. For a similar transcription of a title in Tibetan script at the foot
of a document we may note Or. 11252 (36), where the frequent Khotan
title pharsa (found also in the forms pharsavata, pharsata, pharsava,
pJiarssa, in which -vata is Old Iran, -pati) is reproduced in Tibetan
script as pha r(a) sa. In connection with the origin of these documents
of Sudarrjain, it may be noted that besides other places phi mail a kltha
(loc. sing.) occurs. This is the phimdmna Jcamtha of the Sacu document
(ed. Konow), Chinese 0 p'iei-mud, Tib. phye-ma (F. W. Thomas,
loc. cit., p. 24).
In these documents of Sudarrjam we have the dates
Or. 11344 (3) B 10 : 33 ksuni.
Or. 11252 (30) 1 : 35 ksuni.
Or. 11252 (34) 7 ; sala kseradirsa ‘ year 36 '.
Or. 11252 (16) B3: sparadirsamye ksa ‘thirty-fifth ksuna'.
Since the General Sudarrjam is mentioned in each document, we
can evidently refer them all to a king who reigned thirty-six years,
and, since the Tibetan script is used, subsequently to the Tibetan
conquest of Khotan.
’ This use of o = a suggests an explanation of the word, ibidem, p. 116, node
' dramatic performance ’intheLi(= Khotan) countrj'. It would correspond to Khotan
Saka *7talai < Prakrit nadaga- (Ardha-Magadhi nadaga-, nadaya- ‘ drama ') < Skt.
nataka-, cf. F. W. Thomas, JRAS., 192.5, 498 If. The word is attested in Central Asia
in Dialect A [= Agnean] natkam loc. sing., and in Kuchean nafak. In the Dictionary
of Dirghayur-indrajina (che-riT3-dba73-rgyal), od. Bacot 118 6 I, nalaka is rendered by
bro-gar ‘ drama ’. In no-le, no = tia and le = lai. For -e = Khotan Saka -ai we hav-e
a second example in Tib. phyi-se, phye-se, ibidem, p. 25, Khotan Saka pisai ‘ teacher ’,
in which I prefer to see Old Iran. *patidaisaka- to dais- ‘ to show ’, Mid. Parth.
'bdystn pres, 'bdys- ' to show, teach Oss. dicdesun ' to show, inform ', fades ' crx' of
alarm ’. This may also be the place to indicate a better etymology of Khotan Saka
pir-, ptc. pida- ‘ to uTite ’ as from Old Iran, pnti-kar- ' to imitate, copy ’, used in Old
Persian of the rock sculptures of Behistun, and in Mid. Pers. patkar, NPers. pnilnr.
Armen, patker for ‘ representation, picture ’. Cf. also Mid. Pers. nikdrak ‘ a diagram ',
KPers. nigdr, nigdstan. Hence we have *patidai- > pi and *patika- > pi-.
936
HVATANICA
In other documents three royal names are found : —
(1) Visa dharma.
Ch. 1,0021a, a 20, \dsa darma baudasatvam ramdi ksuni ye puha
cu hija kamala asiri basti sail haudama masci ‘ Of Vim Dharma, the
Bodhisattra, the King, it was (ye = vye) the fifth ksuna, the year when
the Red Heads lerought evil, the seventh month.’
The hija kamala ‘ Red Heads ’ may refer to Tibetans, who are
named ‘ Red Faces ' in Tibetan texts, cf. F. W. Thomas, loc. cit..
p. 78. We know the Tibetans also as ‘ Black Heads ’, mgo nag (JRAS..
1910, 955).
(2) Visa sura (not hitherto noted in Tibetan or Khotan texts).
Ch. 1, 00216, a 43, puhye ksuamni pasa silya ttaujiinri masti
harndusainmye hadai .... visa sfiri ‘fifth regnal year. Sheep year, the
month of Ttamjara, the seventeenth day, . . . of Vim Sura
Ch. 00274, 2 r 4, the Jataka-stava,^ in the introduction : M v%
surra mistye rrumda hye udisaya ‘for the welfare (hita-) of the Great
King Sri Vim sura
(3) Visa Kirti.
M. Tagh c. 0018 : visa kirtta (facsimile in Serindia, plate cli).
il. Tagh b ii, 0065 (facsimile ibidem) ksasimi starn tsve ksumni
‘ his sixteenth ksuna passed ' in a poem celebrating Visa kirtta of
hvam ksiri ‘ Khotan ’.
We have therefore the following chronological data : —
Vi& sura 5 + a; years beginning in a Hare year.
Visa dharnra 5 -f a; years.
Ahsa kirti 16 + a: years.
Reigns of unknown kings : —
36 -f X years after the Tibetan conquest.
24 -r a- years, beginning in a Sheep year.
30 ~ X years, .. ., Monkey year.
9 -j- a- years, .. ., Hog year.
14 4- a; years, ,, „ Dog year.
' It ha.s not yet been possible to learn if the Jataka-stava of the Derge Tanjur
(Tohoku Catalogue, no. 1178) is the same or a similar poem.
Old-Iranian “ Peership ”
By Ernst Herzfeld
TN the course of studies connected with the social order of Iranian
antiquity, I have broached, without expounding it in context,
a problem contained in the expressions vispatis, visa pii0ra, and
vuxpuhr. The matter has been taken up by H. H. Schaeder in his con-
tribution to the volume dedicated to Sir George Grierson “ Ein
parthischer Titel im Soghdischen,” where, while adopting the main
results which I had abstracted from my epigraphical material, he
argues against opinions imputed to, but never put forth by me. I may,
therefore, be allowed to explain the view I really hold.
Back to prehistoric times goes the fourfold graduation of Iranian
social order into nrndna- “ house vis- “ clan ”, zantu- “ tribe ”, and
(lahyu- “ people ”. Each term may signify the people themselves as
well as their habitat. Each degree has its chief. (1) nmdndpatis.
Germ. hausherr ”, is etymologically the same as Gr. SemroTT;? ;
(2) vispatis, chief of the vicus “ clan ”, or paterfamilias in the oldest
acceptance of that Latin term, comes near to Engl. “ earl ” ; (3)
zantupatis, rendered in later Greek (Agathangelos) by yevedpxrjs,
since zantu- is a large subdivision of, or sometimes a satrapy, assumes
the meaning almost of comes, dux ; (4) dahyupatis, Gr. iOvdpxys, can
be the satrap, but in Media where the old empire coincided with the
later Persian satrapy, it was the sovereign title, which it remained under
changed pohtical conditions down to the time of the Paikuli inscription,
1. 31 ; dryan xsatr masist xvatdvi u dehpati “ of Eransahr the most
great lord and sovereign ", and down to al-Berunl, who explains
al-dahufadhiyya “ sovereignty ” by “ the office of guarding the world
and ruling over it ".
The various chiefs and their nmdym- " kleinfamilie ", constitute
the high nobility which ranges above the class of the tunvant-, Alck.
mdr-bam, the “ vollfreie ”. The higher degrees of zantupatis and
dahyupatis mean greater power, but no social distinction : they were
all vispatis. A. Christensen is perfectly right in calling the chiefs of
the great families of the Achaemenian epoch vispatiL There can be
no distinction between “ die Clan-herrn schlechthin ” and ” the chiefs
of an aristocratic family ”A for, to be chief of a clan is the only title
to aristocracy.
^ Sch., I.C., p. 747.
938
E, HERZFELD —
Those families, of which a great number are mentioned in Grei
authors, as well as in the Awesta. were called after an eponyniii
Some of the families may be traced through almost a thousand year ■
The oldest is the house of Deiokes. Sargon of Assur mentions tL
eponymus in 715 B.c. as governor of the province of Man, under tl.'
King of Urartu. His descendants, the Deiocids of Herodotus, becarn
the Median great-kings, and we can follow their career from chief -
of their clan to sovereignty. The Spitama, the family of Zarathustra
appear in Ctesias's Median History as the owners of a great part e
Media, which must have been Rhages. Their name is mentioned i'
Babylonian documents of the fifth century. Another house were tie
Vlvahana, iii the Pap. Elefantine Vaivahana, “ of the family of ^ iva.
father of the first man Yama. They are mentioned in Zarathustra '
Gatha Y. 51, 12, where vaiPyo is a disfiguration of vaiVallyo : " N"’
did the royal Vaivahya, in (the month of) jtrto.zimo (December)
welcome him, the Spitama Zarathustra, as he debarred him fron
shelter, when his carriage-and-pair put up with him and the horse-
were shivering with cold.” One of them was the satrap of Arachosi,.
under Darius, who is called only by his family name. Later that
name became M.Med. Vezan or, synonymous, Geivpiihrdn. The}
held the province Hyrcania as hereditary fief, and the Gathic verses
allude to Zarathustra's " hidjra " from his home, Bhages, via
Hyrcania, to Taosa, the residence of his protector Vistaspa.
Another of the grandees of Darius, Gobryas, whose clan according
to Strabon hailed from Persis, is called pdtishvaris in the inscriptions,
i.e. zantnpatis of the country Patishvara, later Padisxv'ar, Tabaristaii.
The fact that the vis is called after its fief proves that they held it
already a long time before Darius.
Such dukes or princes or kings were the most successful among the
vispatis, but being the chief of their clan was the conditio sine qua non}
That the Achaemenids were one among many aristocratic clans needs
not to be inferred from the fact that, in speaking of their own family
or its residence, they use the expression this vis ” ; they use that
expression because they are the vis of the eponymus Achaemenes.
Under their rule the chiefs of the leading houses had the
right of unannounced entrance to the king — similar to the grandees of
^ I cannot refrain from telling a rather pertinent anecdote I remember from my
childhood ; a member of the Quitzow family said to a Hohenzollern prince : “
have been in the Mark long before the Hohenzollern.’’ The prince answered : Perhaps
we were later, but surely more successful,’*
OLD-IBAXIAN PEERSHIP "
939
Spain and the ambassadors of old — of saluting the king by a kiss, of
sharing his meals, and of intermarriage. That shows that they were
peers. Herodotus explains some of those prerogatives as agreed upon
between Darius and his six helpers at the assassination of the magus
Gaumata ; but it stands to reason that those are old inherited customs.
The language must have had a term for the peership, which I recognize
in the Awestic word xvaetus, rendered into Greek by avyyevels}
Already Darius tried to restrict these prerogatives, and historical
developments naturally lead to a limitation of the number of such
privileged chiefs of clans.
The dignity of vispatis probably belonged to the eldest member of
the family, in the case of the later royal houses not necessarily to the
king. From this “ chief of the house ’’ vispatis, the term visa piiOra
must be distinguished.^ It means literally “ son better “ heir of
the clan Its full meaning is " heir to the chieftainship of the clan ".
then, a little wider, “ born, member of the chief family of a clan ;
but it never includes the clients, the feudal vassals of the chiefs. Such
is the meaning from the beginning and always.
In the Awesta visa puOra- occurs only twice. First in Yt., v, as an
epithet of the hero ©raitauna, the visa pudro adwyanis, that means
heir ”, in later language vdspuhr of the clan Adivya not a vague
and insignificant “ noble scion The other instance is Yidevdad, vii.
43 : a physician is entitled to as much fees from a visa pudra- as from
a zantupatis ; a dahyupatis only has to pay more, a vispatis much
less. As the “ son of the clan ” here ranks above the “ chief of the
clan" with the zantupatis = duke, only second to the dahyupatis
— sovereign, obviously the term is used with a latent qualification ;
it is the son of the royal clan, the clan Kan' i^oxyv. an expression
I used in AMI., ii, 32. and vii, 18. and Schaeder twice on p. 743. That
is the reason why in the OP. inscriptions viO- stands for the royal clan
and for their residences, Persepolis and Agbatana. In the same way,
the residence of Yistaspa, the town *Taosa naotardnam. later Tos i
Nodar, is called vis naotardnam in 17., xv,^
' Ristovtz’lf accepts this view in C.ambr. Aiic. Hist. Ill, iii, p. 114, and identifies
them also with the megistanes of Josephus.
AMI., IV. 34.
" AMI., ii. 52.
* AMI., vi. 74. But the reading viBbU in Bar. Pers.e, three times, is wrong ; the
text has thrice the normal hada visaibis bagaibil “ with all the gods ” ; there were no
gods of the clans.”
940
E. HERZFELD
The special meaning of riso jmdra in the Yidevdad cannot be of if
than the establishment of the kingdom in Iran. But it is nothing tin
an arbitrary assumption that visa pu6ra, at that period, was soh i
used for a son of the royal clan. The expression inso pudro ddu'ijdii
has certainly not been introduced into the legend of ©raitauna a
late as the Achaemenian empire : such expressions are primary. An
its unqualified meaning, heir to the chieftainship of the clan, wln i.
applied to the royal clan, by itself becomes heir to the throne
Nor was the application of vi6- in OP. limited to the royal clan, b
NiR h, Darius addresses the reader utd vidiyd utd spadmaidayd, Akk
ina hiti u ina madaktum “ whether in a clan or in the army ‘
prehistoric expression, for it recurs in Latin dorni militiaeque ; there )'
no question of the royal clan ; he does not speak to princes, but ti
warriors. In a similar way, in Beh. § 14, vid- is opposed to kdra. Akk
uqu, i.e. the militarj', lower nobibty, the “ vollfreien, wehradel
The Elamite had not the ideogram UL-HI, normal translation fin
vi6- in the sense of royal clan, residence, but apparently ulamnumi"
with the ending of a personal plural. The passage refers to two classf-'-
of the nobility but not to members of the royal house.
Another instance is the expression kdra hya viddpatiy in Beh.
§§ 24 and 40. The translation, “ the army stationed at or near the
vid-, the guards " ^ implies again the limitation of the meaning of
vid- to the royal house. Darius speaks of a body of troops, in 24 in
the province of Mada, in 40 of Parsa, that mutinied in his absence.
In the first case, the king was somewhere between Parsa and Elam,
in the second, engaged in battles in Bactria and other provinces. He
would scarcely have left, in his precarious situation, his own troops idle
in Parsa or Agbatana. The expression must have a very special mean-
ing, for the embarrassed translators do not reach a uniform rendering.
The Elamite avoids, as in § 14 , the ideogram UL-HI, and uses appo
ulammanni those in the U ”. The Akkadian has, in 24, mala
ina hlti all that in the B.” (blla means house, clan, also fief), but in
40, mdla ina alluka sa dli. dlum “ village ” is written with the ideogram
ER. as in the Assyrian rendering of Med. vispatis by EN.ER. ne-
bel-dli “chief of a village, clan’’, alluka belongs to ilku "statute
labour, Fron ”. The phrase, hence, means “ all those in the feudal
service of the clans “. The OP. version says that the troops hacd
Y addyd fratarta “ secessi ex Y.’’ The reading or the writing of Ydaya
‘ Sch., I.C., p. 743, n. 2.
OLD-IRANIAX " PEERSHIP
941
must be erroneous. Before, I felt inclined to restore Hdaya, after Skr.
sahdya " auxiliary But ha\'ing discussed the problem with
A. Yahuda, I prefer a synonym of ilku — allitka, the OP. rendering of
Akk. mandattu, Aram, mandci, hence Mdaya. hacd nmnddyci fratarta
means ” they deserted from their duty of service
The § 40 thus decides the problem : Idra hya vidapativ cannot mean
the royal guards which were ” soldiers ”, Gr. jxiaOo^opoi, Oir.
*ndzddvara in the spddmaidayd, but only the troops provided by the
high nobihty, called avpLpa)(OL by Xenophon, a militia or yeomanry
of the clans — vidiyd — of Persis or Media, viddpativ, a compound of
vid- with the postposition paGt', is what Herodotus expresses bv " thev
lived Kara Koyias, kw/jlt] being the vid- as a place ; as clan it would
be (j^parpa. viddpati» is exactly /card (ftparpa? ” organized in
clans ”, in contradistinction to the otherwise organized troops of the
satraps and the king. The adverbial compound of substantive plus
postposition has the value of an adjective, like its entire type which
survives in MP. and NP. Neither in Awestic nor in OP. has vid- ever
ceased to signify the clan in general.
The MMed. or MParth. form of visa pudra is vispuhr, from which is
derived, with vrddki, as already explained by Bartholomae in IFZA’3/.,
25. 1911, pp. 251—4, the adjective vdspuhr, i.e. ” vispuhr-iau In
the Sasanian inscriptions, whether Pahlavik or Parsik. the word is
always written with the ideogram brbyt' ” son of the clan ”, for which
the reading vispuhr had been generally adopted, also by mvself in the
Paikuli book. A secondary adjectival derivation in -akdn occurs in
phonetical spelling in the Naqsh i Rustam inscription, ^ which, from the
old photograph of Stolze, I had read, as Noldeke did,- vUpuhrakdn.
After having discovered, on the rock, that vdspuhrakdn was the right
spelling, I annotated in AMI. vii, 19 : “ [fn.spw^raA'na] : Paikuli, p. 170,
und wo sonst dies wort vorkommt, dessen simple.x vdspuhr immer
ideographisch geschrieben ist, ist danach tm- fiir I'G- zu verbessern,”
That means, without ambiguity, the corrections to be carried out in
my Paikuh book. In AMI., ii, 32, I remarked that " die rangklasse
des hochadels in arsakidischer und sasanidischer zeit vdspuhr. nicht
vispuhr genannt wiirde.” tVhile adopting both statements as " ohne
zweifel richtig ", Schaeder, quoting the two places, strongly objects to
* I profit of the occasion to correct myself : m ,4, Iff., vii, p. 19, 1. 8, the abridged
names Sar Mashhad and Xaqsh i Rustam for the two inscription.^, arc by mistake
exchanged ; the passage containing the word taspuhrakan stands in yiRn., and is
omitted in HMshh.
^ 1 had overlooked the reading with a m Westergaard’s rare book.
942
E. HERZFELD
my contesting the form and denying the independent subsi^tt ;, "
of rispiihr for the Arsacidan and Sasanian time in general Agai' 't
this undeserved accusation I plead Not Guilty ”, although it wi ’; 1
be totally indifferent.
It had been Bartholomae's opinion that vdspiifir signified ‘ e
members of the Iranian nobility only “ inasmuch as they were, r
were believed to be, descendants of the royal house ". Schaedt ;
slight modification, that it originally meant the descendants of t' ■
royal princes, and later became transferred to their equals, the menil „• -
of the highest aristocracy, is based on the presumption that a cl i
distinction existed between the simple tnspithr, the “ prince royal
and the adjectival derivation vdspuhr “ which no longer signified t : ,
royal princes but the members of the high nobility ”. According ’ '
this theory, OIr. vi-S puOra-, though originally meaning “ son ot .
clan", was never used as such, but solely for the heir to the throii'
in the reverse, vdspuhr, though derived from the word for the " h' "
to the throne ”, was never used as such, but exclusively for t
members of the aristocracy. Both contrary assumptions are erroneou".
In the Paikuli inscription, two royal princes, Peroc and Narsahe- -
both, if I remember rightly, uncles of the King Narsahe — are twn'i
called brbyt', vdspuhr e sdsdnakdn “the Sasanian prince royal
Hence, vdspuhr, not vispuhr, was the predicate of the royal prince'
and at the same time of the high noblesse. In the latter meaning n
appears in the series sahrddrdn, vdspuhrdn, vazurkdn, dzdtdn, the social
signification of which cannot be questioned ^ as it not only occurs in
Hajjiabad but several times in Paikuli in a long context. The
distinction at the Partho-Sasanian period between vispuhr " prince
and vdspuhr “ peer ” is not ascertained,^ but imaginary and con-
tradicted by epigraphical evidence. Exactly as OIr. visa pu9ra-,
thus MP. vdspuhr has both the acceptances.
The application to the high nobility is well known. But the
“ erweislich unzutreffende These " of the application to the prince
royal is worth studying. Since the fundamental idea is the son as the
heir at law, “ son ” alone can have the value of a title when used of the
son of a ruling king. Already in Assyrian, mdr sarri " king's son is
' When writing the commentary to Paikuli in 1921, I had not yet studied the
question, whether the two princes might have been “ heir presumptives ”, and I
thought at that time that one or both of them might have been vispatis of
the roy.'il clan.
2 Sch., I.C., p. 744, n. 1.
’ Sch., I.C., p. 746.
OLD-IKANIAN “ PEERSHIP ”
943
a princely title, and mar sarn rahu “ the king's great son ” is that of the
heir to the throne. The same in Old Iranian. In Yt., xiv, 59, the simple
pudrdho “the sons”, the TratSes- of Thukydides, is a title either of
the royal princes or of the cadets of the high nobility. The OP. text
of a trihngual inscription in the tacara of Darius at Persepolis, much
mutilated, may be restored to “ Xerxes, the son of King Darius, the
Achsemenid ”. That is Xerxes as heir apparent. As the Harem-
inscription of Xerxes shows, maOista, said of a son of the king, means
the designated heir to the throne. Likewise, in the last verse of the
last Gatha, F., 53, 8-9, Zarathustra prophesies to the one that kills
the “ Evildoer i.e. the magus Gaumata, that he soon shall be
mazista, i.e. shall inherit the throne.
In the Iranian epics the King Godarz, i.e. Gotarzes II, has a son
Vezan, from OMed. *vaivazana, OP. Vaimhana (Pap. Elef.). He is not
an historical figure, but the personified name of the clan, for which the
king in his inscription uses the synonymous form FeoTTodpos, the
Geu'puhr. The family name has been materialized into the son of
Godarz, because, just as 0raitauna is called viso piidro ddwydnis, thus
the heir of Godarz bore the title vaspuhr e vezan, OMed. *ris5 pudro
vaivazanis.
In the Sur Afrin, a late Sasanian pamphlet, the heir to the throne
is addressed, after the king, and before the vazurg framdtar (while in
Paikuli the two vaspuhr e sasanakdn follow the hargupet or grand-
vizier), with the words pus i vaspuhr i sdhdn farraxvtum, etc. In AMI.,
11 , 20, n. 1, I had translated “ vdspuhr-Sohn {wali-akd) ” and in vii,
18, “ infant, sohn des (kbnigUchen) hauses, gliickseligster (or glor-
reichster) der konige ”. pus alone, in a ceremonial address to the
prince royal, is certainly not a pet-name, hence, justifies " infant " ;
vaspuhr and sdhdn farraxvtum stand, hence may be translated, in
parallelism ; all royal princes are “ kings ” ; '' vdspuhr-soim ” means
'■ vispuhrischer sohn " ; Schaeder translates " hochadliger Sohn ",
that is exactly the same, but censures my translation as “ unhaltbar ”,
as if it became “ tenable ’’ by repetition. Those are trifles. But an
important point is : by no means is vaspuhr an “ epitheton ornans "
in the sense of an arbitrary epithet. All such titles are ornans, and yet
they constitute the official protocol. Unconditionally we must allow
vdspuhr instead of vispuhr to be the official predicate of the heir
apparent.
The adjectivml derivation vdspuhrakdn. on the phonetic spelling of
which in NiRst. the reading vdspuhr of the ideogram hrhyt' is based,
944
E. HERZFELD
occurs in the title of a high official, vasjnirakan hanmrkar in the hi't v
of the Armenian Sebeos. From comparison with the v’arious el i '
of titles formed with local appositions, and especially with oti il
seals of the amdrkars of Garamaea and Mosul, to which I could ! '■
joined Ears, I had ventured, against the authority of Hubschmaii!' n
explain vaspurakan hamarkar as “tax-collector of (the district i)
Vaspurakan This is another of my finds which Schaeder acre]
adding as confirmation the alurpdiakan arnarkar of the Dari i:. 1
inscriptions. If one concedes that vaspuhrakdn in this connection caii! f
mean “ of the high nobihty ”, but must be the name of a district.
necessarily is Isfahan, for the theatre of Sebeos' story is Isfahan, a : i
it is expressly said that the collector had the taxes of Isfahan in 1 '
house. The identification is quite independent from the meaning t
vaspuhr. Adducing the correlative evidence of the “ Royal Sopheiu
and the “ Royal Marw ”, called thus for having been under the din t
administration of the crown, I explained the name vaspuhrakfin . -
“ under the governorship of the vaspuhr. the crown-prince ". " I' ’
the same reason,” I wrote, “ the Armenian province (of Vaspurakai
received its name.” This scarcely justifies the response “ Keinestal •
kann die armenisch-persische Grenzprovinz V. im Osten des Wan-So -
das Stammland der Arcrunier gemeint sein I never identified th
Vaspuhrakan of the inscription with the Armenian province. Tin
reasons alleged against the identification speak in favour of tin
explanation of the name ; Vaspurakan is a frontier province and if"-
name can probably be traced back to a period of Armenian emancipa-
tion, in the second century b.c. There are plenty of examples — from
the Assyrian period on — of the heir to a throne preparing himself for
his future task by administering a frontier province, e.g. the Sasaiiian
prince-governors first in Bactria, then in Sakastan ; and, of course,
that could only have been at a period of independence, e.g. in the
second century b.c. Sebeos. under the name of Vaspurakan.
undoubtedly speaks of Isfahan ; the Karter-inscription of IsiRst.
contains no clue as to the locality, but the Armenian province is almost
out of question, and it would be arbitrary to assume a third
Vaspuhrakan. Isfahan must have received that official name, because
it had been, probably during the second half of the Arsacidan period,
between a.d. 50 and a.d. 225, under the administration of the
vaspuhr, the crown-prince. Under the Qajars Tabriz was the province
of the wali-ahd.
As to the merely philological side of the problem, we must state
945
OLD-IEANIAN “ PEERSHIP ”
that, although vispuJir would be the primary designation for the heir
to the throne, actually, during the Sasanian period, the Arsacidan
word vdspiihr was employed. The two words are to each other as
e.g. consul to consularis, praetor to praetorius. The derived adjective
has been used as a substantive and has replaced the original. Since
the notion of vaspuhr includes the vispuhr, the substitution is not
against logics : the vaspuhr hat'exochen is the vispuhr. Distinctions
like that between a primary vispuhr and a secondary vaspuhr may be
grammatically right, but are unfit for proving historical conclusions.
For the application of such closely related words is not ruled by con-
siderations of grammar, but by practice. Which one of them is
eventually used and which is its specific acceptance at a defined period
is not a question of philology but of history.
A New Contribution to the Materials concerning the
Life of Zoroaster
By S. H. Taqizabeh
rpHE life of Zoroaster has been the subject of both comprehensive
and exhaustive study since Hyde and Ancjuetil. The amount
written on this difficult question and the discussions which have taken
place among scholars are disproportionate to the insufficient and
fragmentary materials which we possess. Even the available ilata
are lacking in precision and are not very reliable. However, every old
document available relating to Zoroaster ha.s been carefully examined
and most of them collected by students of Iranian history. The out-
standing collection of these scattered fragments, so far. is the famous
work of Professor A. V. W. Jackson. Zoroaster, the Prophet of Ancient
Iran, to which .some subsequent additions have been made by the
author himself, his pupils, and others.
A new note relating to the life of the Iranian prophet written by
Biruni. the famous and eminent man of learning of the eleventh
century, has recently come to light, which, no doubt, deserves to be
added to the materials already published. This is an article in Biruni's
well-known book Al-Atliar al-baqiyat 'an il garun al-Khaliyat. better
known as Chronohgij of Ancient yations. edited in Araljic text and
translated into English bv Sachau. The text of this edition, however,
has unfort unatelv manv lacunae owing to the defectiveness of the
original manuscripts on which it was ba.sed. In those copies which
nere at Hachau's disposal large portions of the text had in many
instances been omitted. One of these lacmue happens to be in the
ciiapter dealing with the sulqect of heretics and the founders of false
religions, which is covered bv pages 204 -214 of Sachau's edition. It
is <>vident that a suiistantial part is mi.ssing from the end of page 20.').
which fact .Sachau himself indicated in the footnote with the words
'■ Grosse Liicke ",
In Teheran (Persia) there are some copies of this hook which were
not known to , Sachau. One of these is preserved in the Library of
Sepahsalar'.s Mosque (now converted into the School of Theology)
and perhajis another was known to that learned Prince. All-qoll
INHrza E tklad es-Saltaneh. who was the Persian Minister for Public
Education about half a centurv ago. and who gives extracts from it in
viiL. vm. p.\Kr 4.
01
948
S. H. TAQIZADEH —
his learned book Falak es-sa‘adat, printed in Teheran in a ii
which extracts are at v'ariance with Sachau's edition.
Now in a new and interesting book entitled Ddnishni'U"^'
Azarbdyijdn and compiled by Muhammad-‘Ali Tarbivat (jf T.;
the author, who following some traditions believ'ed that Zort
was a native of Azarbayijan, devotes an article to this ” most fo'
man of that province The book in question has just been puhli'
in Teheran and comprises 414 pages, and is, in fact, a biograp! -
dictionary of the learned men of the province of Azarbayijan (ii
wider sense of the word) arranged alphabetically. In the ai’
Zardosht , pp. 163-4, the author, after writing a few lines about ■
modern researches into the life of Zoroaster, finds the best wat
giving a concise account of the career of the Persian prophet bv quot ■
the text of Biruni s note in his ‘‘ chronology ’ taken apparentlv 1"
one of the above-mentioned copies in Teheran. This note, which cor
•sponds to some of the lacunae in Sachau’s edition, runs as follows .
J"
JUfit dll;
Ljjl _j^\i vIjC tiUll
^ J. J ^ Ll-1 <UiLZ ‘
Ji oA._, Aj Jl jJ ^
Ale ^ Jj; 4; I ^
<>• 1:: — * dCly tJuJl <vl (31:1,9
)_ -tlj 4 III jLj:_V( A-^> \ (Jl
^ t t
A^^Vl aa9 CjI-^VI ^I^’l3.l9
\ (js' p.«5 jl 4 ; 1
THE LIFE OF ZOROASTER
949
JS- j ' J*'
yi>\^\ \ ^1 a 1.^‘ i ^jl^3 a"^! ^ (_r'^^
*0 '"^S/1 j*-5; ‘^'«L';y^ ol'Vl
<«L' _yt> J.. C)l-»!i! <i3lji <«! L':_^'_'
[s'c] ;)\J olJJi j>-
(Ji^Ll-.; (j\; _; jLJ J3> jl— 5 J^' ‘^-’.
Uili A' lil .b f»-«^ cSCL: jUi
• • 1 (^ •
J1 A) j'^ j' J^-5
^ \s aAc- aAai ^1 *!<' » ^r^'« ^
« j_^« ^y-* Oj*. ^ i}^\ (3^**-? ^^3 Ai^ A»ia) .J ojX^
«fe>
aU cU^ <A lii j' J tlr* ^Xj
Y« ^S3'%* ji ^j5 aWIs 3
O''
iUU J_yJ (j' od^L'3 tlr*^’- j' ^ a_IjU As' Xc
A_"^ 3’ > A1)J> d>A»i a
j_r-U^ eX« ^1 y^ -Xs^ (j^*-’. 3^3
^^1-3 3^3"' a" Ar*
XjXs- 3^ ^ XJ-Vs- <Jjji-^ 3
’ ^1 ajiparently.
“ A x\ ord or two st-cms to tx* ml^^iIl^ from here, jn'i'liaf'^ ''i\c must read
Read 0)1
950
S. H. TAQIZADEH
i'j' j <.bi J
W" J-'^^ ^ Jli_ i^j 4i— ^ *
a 1 >X*ii3 Al^jt jL<a Alls I <li j) .’
*
tilj “Cc /«‘:--' “V" ci-^-' y
(_^ A.i I j I _/. -A^l CjULIIs i
"^^Is J.'L Al'lj ^ LT^ij iijjji-i-y '
4.3-1 j jU_^ t_A- — dTj J< '-^-“■>L'J'. ^}j=''
(j?^ «-l— ‘iljj jJ.^4 ^ ^ tZjl«L."»is t_^-i-ll aIj
a; I (1jA3.1 U c1)4>.I j Ajbiji) aJ
A;1-_5^^ All Ale L« Ia* (JJ 4
^^^1 J jl yj- ^1 A.>i 1 ^ jl^ 4J1 jia'
j^'L. I (j jl ^jiji)i (3 jTi As _J AI.4 4 »j1~~--9
«• *» ?
cX Aillii A.«l t %b ^ Vl
u. Ij.j^i'^ jx jx Alii.^j l-«
ajA.M3 3^C- )1 a« (_;-l^ <_Jja.i ^1 jl< ‘^' ^
jlAi_^l C^^ljj iVjl ^ aiV^l J Jbl-j- o_^l
li—i I 3 ^ 3^ (3^^’ -\.' J-*
(S-«b^ Ale O’V (»■«“*
V^jl <3^^^ t>« ‘'e a) *— jU*; V 1 v-Ali 3
THE LIFE OF ZOROASTER
951
1 jii jj". ^ a1
t-
^.AlC VI <s^>.l9 ^'1 :sj^U ^4 jd:^ c^JJl
^ t.— d-\l_5 d-Xi o' ^**1^ (J1-J^*1
t ^ '
jVl (3 jUlj 1^) J;-li'‘ j^Al’ls (_;-li-l ^lll'‘ jA» «o
<1 <— 1^) _J l^ll.) c- ^*' 1 jAs
ItL-l j
l^" Aulfi-S
[!rTfl/(s/rti(o/(]
■' Afterwards came Zaradosht (Zoroaster)d son of f?tidtoman
(Spitaman) of Azarbayijan. who was a descendant of King Minutshehr
and of the Houses of Huqan Vielonging to the nobles and grandees
of that place. This happened when thirty years had elapsed from the
[beginning] of the reign of Vishtasp. He came forth clothed in a tunic
spht at both sides, right and left, wearing a waist-belt {lunndr) of
palm-tree fibres, as well as a painted - gown of felt and having with him
a dilapidated document which he held in his hand against his brea.st.
The iMagians (m-ajds) believe that he had descended from heaven from
The roof of the Court hall in Balkh (Bactra) at midday and that the
roof was opened for him.® Vishtasp came to him (to Zoroaster) from
his midday rest and Zoroaster called him to Magiani.sm (the
Mazdayasnian religion), to the fullilment of belief in God. to magnify
Him and to reverence Him. to reject the worship of the Devil, to the
obedience to Kings, to the correcting of nature, and to the marriage
with next-of-kin.^ As to marriage with mothers, I have heard Sepahbed
Marzban, son of Rostam.® say that Zoroaster did not establish this
^ The words in parent hesof' tliroughmit the translation arc adtlcd by myself.
" The text has ,-*1^ whuh i.s given in dictionaries as meaning a parti-
< '’loured gown of woo! with &tripi‘s of tigures, but 1 presume is used lierc as meaning
a veil.
^ The text has but I prcMime the correct reading is
* Lit. the nearest of the relations.
^ The well-known author of Marzban Xameh. Biruni often mentioned his name as
the source of his information about things relating to old Persia.
952
S. H. TAQIZADEH
for his followers ^ but approved ^ it when Yishtasp called befon
(Zoroaster) an assembly of the great men of his time and the If.r I
people, and one of the questions they put to him was about the luar : ' >
living with his mother alone, far from other people and having no at '
to women except to her, felt disquiet lest he should have no descend
Zoroaster replied to them that in this ca.se intercourse with the m"i r
was permissible. He brought a book which they call ' Alx-
- (A vesta) ; this is written in a language differing from those of all tc
nations. It is based on a unique idiom which contains letters exceed j
in number the letters of all languages, so that the accjuaintancc i
may not be the monopoly of the people (speakers) of one langu^
to the exclusion of the others. He .set the book before Yishtasp wl i
the learned men of the country were pre.sent and the crowd " -
gathered there. Zoroaster ordered copper to be melted and tii
said ; Oh God, if this is your book with which vou have sent iiif
this King, then prevent the copper harming me.’ Then he ordered r
molten copper to be poured over his bodv ; it was pomcd o^■er I
breast and his abdomen, but the copper ran over and beneath In
[without doing any harm to him] and from each of his hairs hung
round bullet of copper. I have heard that these bullets were preset'M ‘
in their (Persian) Treasury during their sovereignty. Yishtasp accept' '
[the new rehgion] and professed that angels had come to him froiu
God and ordered him to believe oji Zoroaster when he came [and] t''
accept the message he brought. Zoroaster livefl after this and preaclicc
his religion for seventy years, though some say (only) forty-si.x year>
The Hebrews assert that Zoroaster was one of the disciples of tin
prophet Elijah. He (Zoroaster) saifl of himself in the book ot
nativities ^ that he had in his youth acquired knowledge in Harran ‘
from Elbus (?) the philosopher. According to the Romans he was a
* Lit. he (lid not legUlatc thi.'i.
- Permitted.
This book, to n hieh the Muhammadan autliors on Astrology often refer, a-serihiiiL,'
it to Zoroa.ster and .-soraetimes quoting him without giving the title of the book, .scem-
to he an apoorv phal hook attributed to him. The book, which must have contained old
Persian A.strologv, especially that part relating to nativity or honocopes, might have
been the work of some of the Persian Astrologers of the late Babylonian and Syrian
schools in the Sa.s.sanidn period who had the name of Zoroa.ster,' which was not an
uncommon name at that time. The materials derived from the book by Abu M'ash.u
of Balkh and others show also some relation to Greek Astrologj', perhaps coming
through Harran. The oldest Arabian sources refer to this Zoroaster as Zaradusht tin
philosopher
The well-known centre of a pagan community with the Hellenistic culture.
THE LIFE OF ZOROASTER
953
native of Mawsil. They may have included, in this case, the frontiers
(the region) of Azarbayijan in those of Mawsil. As to what the Greeks
•■^aid about him, this is related by Ammonios ^ in his book on the
doctrines of the philosophers. According to this version Pythagoras
liad two disciples, of whom one was called Philayus ^ and the other
Phylacus.® The former went to India, where Brahman, the founder
I if Brahmanism, became his disciple for sev'en years and learned from
him the doctrine of Pythagoras. After the death of Philayus. Brahman
began (to found) his dogma on the Pythagorean doctrine. As to
Phylacus, he went to Babylonia where Wartush,* better known as
Zoroaster, son of Pourkushasp (Pourushaspa) and called Sfidtuman
(Spitaman). met him. The latter learned from Phylacus the doctrine
and after his (Phylacus's) death entered the mountain of Sabalan and
stayed there for some years. There he compiled his book and created
what is well known. But the truth is that he was of Azarbayijan.
wliich fact is proved by what we have related of him that he said [of
himself] in his book of nativities that he used to go with his father to
Harran to meet ® Elbus the philosopher, and to acquire knowledge
from him.
■■ It is registered in the books of history that towards the end of the
reign of Sabfir, the piercer of shoulders.® there appeared a sect adverse
to Zoroastrianism. These [heretical] people were attacked (by argu-
ment) by Adharbad, son of Marsfand, son of Sesb (?), son of Dushrin.
son of iMinutshehr,' and were overcome by him. Then he showed
them a sign (miracle) by ordering melted copper to be poured on his
own breast. This was done, and the copper congealed and did no
harm to him. Then Sapor raised his (Adhardbad's) descendants
together with the descendants of Zoroaster to the rank of
-Mobadanmobads. No one is allowed to have access to the book of
Avcsta which Zoroaster brought, except those who are trusted in
their faith, and whose conduct is found praiseworthy by the followers
^ Fos&ibly Ammonios Heremeias of Alexandria of the fifth century.
’ Possibly Philolaus, the famous I^thagorean and propagator of Pythagoi-eanism
of the fifth century b.c.
^ I am unable to identify this person though it is a common Greek name.
‘ Perhaps a misuriting of Zartush for Zaratushtra corresponding to I^tm Zaratus
<ind Greek Zaratos.
^ The text has but apparently it should be
Dhul-Aktaf, the Sassanian King Shapur II (310-379).
This is the well-known Aturpat Marspandan. As far as I know, this is the only
place where the names of his grandfather, great-grandfather, and great-great-grand-
father are mentioned.
954
THE LIFE OF ZOROASTER
of that religion. Moreover, one can get full access only after a certii '
is written for him wherein it is attested that the leaders of religion !.
authorized access to the book. A copy of this book was kept ii
Treasury of Dara, son of Dara (Darius Codomanus). It was writti n
gold on the skins of twelve oxen. Alexander burned it when he >
demolished the fire temples and killed the Zoroastrian herhe '
(priests). Therefore three-fifths of the book are lost from that ti:
It contained 30 Nasks/ but now what remains in the haini
Zoroastrians is only 12 Nasks. Nask is the name for each section ot '
book similar to asbd‘ ^ by which we name the sections of the Kor.i
PoSTSfRlPTL’.M
Since writing the aVjove article I have been making inquine- -
to the original manuscript of Al-Athar al-baqiyat from which t ■
above-given text quoted in Daneshmandane Azarbayejan was taki ■
The author of this last-named book now informs me that this .r. i
other parts of Biruni's book corresponding to the lacuna of the Sacliai' -
edition were copied by him about twenty-five years ago " from t!
only complete manuscript of the book" preserved in the Sult.c
Bayazid Library of Istanbul (also called " Kutubkhaneye Fniumi
or Public Library). He has kindly sent to me the rest of the fe.\.
wanting, which is about twenty lines preceding the part publisJiti:
by him in Daneshmandane Azarbayejan and some sixty lines (coric-
sponding to the lines of this bulletin) following it (dealing with Zoroaster)-
up to the words in tli'-
last line of page 290 of Sachau's edition. This line, with the following
twelve lines on page 297 of the same edition, constitute the end of tie'
article relating to Zoroaster. I hope to publish these remaining parts
also in the next number of B.S.O.S.
’ TJie fliapter.-! of the Ave.sta.
“ The .-^even .section.-i into which the Koran ii.seci to be divided for convenience.
The Copper Plate Grant of Srivifafaghava Cakravartin
By K. Godavarma
[The following abbreviations of languages and names are U'^lmI in this article: —
languages : B. — Bengali ; G. = Gujarati ; H. — Hindustani ; Kan. — Kaiiarese :
~ Marathi ; Mai. = Malayalam : P. “ Panjabi : Pk. ~ Prakrit ; Tam. = Tamil ;
= Telugii. Names : Han. = Daniel ; Gund. = Gundert ; Jos. = Joseph .
KpI. ^ Kelu Nair ; -Venk. = Venkavva.]
^UHE plate in question has been the subject of great controversy.
^ The facsimiles of the plate have appeared in Epii/raphica I)idica<
vol. iv, and in Indian Antiquary, vol. liii. Translations of the plate
with comments have been published bv Dr. Gundert {Madras Journal
of Literature and Science, vol. xiii. part i. pp. 115-125). Kookel Kelu
Nair (ibid., vol. xxi, pp. 35-8). Venkavva (Epiyraphica Indica. vol. iv.
pp. 290-7). and Mr. K. N. Daniel {Indian Antiquary, vol. liii. pp. 185-
190. 219-229. 244-251. and 257-261).
Tlie fact that the plate has attracted the attention of many
liistorians. linguists, aiul palaeographists is itself ample proof for its
importance from various points of view. In this article I propose to
give a true transliteration of the document and to discuss the
etymology of the name of the donee, the significance of the term
Mnnillufdniapattuni . the formation and meaning of certain expressions
regarding which the opinions of scholars are seen to differ, the linguistic
peculiarities worthy of note in the copper plate, and the probable date
of the grant.
I
TraXSLITER.XTIOX 1
Han no inahUyampate naina srlpilpdlanarapati srlvlralefa
iasakrarartti dtiyayi inurainuraiyc paJa ndrdyirattdntu
I'cnhil natattdyi ninta srlnfardyliava-iakracarttilku tiruvtrd
cyatn. cclldyi ninta makarattnl riydiain udnandyaru irupattontu
ccnta canirdhanindl pcfuiikdyitakuttirnnnaruJa tnakvtaiyar pattina
ttu I rarikdrtta na ndya ccra indnidkappcru n -cettjkku um n t kki fd ma
puttaiikututtoni cildvdteyuin pavanattdiikum peruperum katuttu
vaJenciyainu ralanciyattil taniccettum murccollum munna
* The bvsteiu of the Royal A>iatic Society is folloned for transliteration, except
a, «. r, r, I, and », where (( is u.sed for nn. « for (-«-). r for o. r for n. . for 'f ,
and ti for the hnal attenuated ii of Malayalam.
956
K. GODAVARMA
teyum fancavadyamum sankhum pakalvilakkum puvatayum awi' -
Jcotta
kkutayum vatukapparayum itupalitdfanamum ndlucefikkum tui'-
ccettum kututtom vdniyafum aihkammalafeyiim atima kiduttnn
nakafattukku karttdvdya ifavikdrttanukku parakontalmnn nimk
tu tukki niil kontu pdki ennintatilmn etukkmtatiluinuppi
notu sarkkafayotu kastufiyotu vilakkennayotu itayil uUata eppi
ppattatinum tafakum atinatutta eiinkarnum kiita kotuiikiildr a[i'
yotu kopufattotu I'isesdl ndlutaliyum talikkatutta kifdm-atlotjhi
yil nlrmutalayi ceppetu eluti kiifiittom cefamanlokappefiti'i
cettiydna ifavikorttanukku ivan makkalmakkalkke vahvahiji
kakkutu
ttom itariyum paniiyiir kifdmamu cdkirakkirdmamitni anyakkut’^:
vendtum dtundtumariyakkututtdm erdndtum ralluvandlt()»anyalh
tom candrd
dityakahdla ndlekku kututtdm ivarkalariya ceppeleluliya ceidinHi'l'
pefuntat
tan nampi cateyan kaiyehiUu.
ir
Donee
The donee is referred to in the grant by the name of IfavikorttuJ!
in one place and Ifavikorttan in two places. Of these two form.-'. ’ ' '
possibility of Ifavikorttahan being the original and Ifavikort^
being a corrupt form of the same, has not been considered by tb'
who have commented upon the grant hitherto. Even with regard '
this Ifavikorttan, two opinions have gained ground, namely tb
Ifavikorttan is a non-Christian and that he is a Christian. Dr. Oundei >
Kookel Kelu Nair. and Venkayya opine that the name is non-Christi.i i
in spirit (ref. Madras Journal, vol. xiii, part i. pp. 120 and 146 : ibid
vol. xxi, p. 40 ; Epigraphica Indica, vol. iv, p. 292). In recent yeai-
Mr. K. Is'. Daniel has made an attempt to show that the donee mu'-i
have belonged to the Syrian Christian community. According to hni
the name Ifavikorttahan is a mistake. Korttan, he thinks, is probablv
deriv'ed from Karttan, i.e. ” Lord ". In this connection he also draw '
attention to the clergymen of the Christian community being calleii
Kattahars. Kattahar. in his opinion, is no doubt Karttahar. which i'
the honorific form of Karttan.
I cannot agree with Mr. Daniel, inasmuch as the arguments he ha'
advanced to prove his point (see Dissertations an the Copper Platen
COPPER PLATE GRANT OF SRIVIRARAGHAVA CAKRAVARTXN 957
in possession of the St. Thomas Christians, p. 5) are not at all convincing,
and the etymology he suggests is phonologically faulty. The change of
Karttan to Korttan has not been substantiated by examples ; nor
has the principles underlying the change been enunciated by him.
Of the two forms Korttahan and Korttan. I am for considering
Ivorttahan to be the original. Korttahan can change by the loss of the
tinal syllable to Korttan. On the other hand, Korttan can nev'er assume
the form Korttahan with the addition of a new syllable without any
■significance whatever.
The next point to be considered is the form that would admit of
giving rise to the one we meet with in the copper plate. It is to be
remarked here that there is only one script employed to denote long
and short o throughout the plate. \\ hat has been read as
Korttahan or Korttan must have been really Korttahan. which can
only be a Tadbhava of Sk. Govardhana— with the elision of ra
after o. That such disappearance of ra after o. long or short, is a
common phonological phenomenon in Malayalam. is borne out by
examples as aivatii, covatu, cdtii " foot or the underpart " ; ciivappu.
ronippu. coppu “■ red colour '' ; tuiarnn, tovofan. tofnn " fried
vegetable dish” ; kovani. koni "ladder " ; kovanani. k6na)n. etc.
It is the light contact involved in the pronunciation of va coupled
with its following a vowel having the same place of articulation as
that recjuired for the utterance of va, that causes the elision of the
latter. Names like Gordhan Das (Govardhana Dasa) obtaining
m Northern India exhibiting a similar change also go to corroborate
the above interpretation. The appearance of voiced stops of other
languages as voiceless stops in the loanwords of IMalayalam need not
be illustrated here, as it is a very familiar sound change in Malayalam.
t before favi is a prothetic vowel which is found to develop before r
and 1 in early Indo- Aryan borrowings. Cf. Iranian for Sk. Rama- :
ilavahnam for Sk. lavahga-. Thus Ifavikorttahan is obviously a
natural and scientific phonological modification of Ravigo-vardhana.
which undoubtedly is a Hindu name.
As to the identitv of the Ravigovardhana on whom was conferred
the Manikkifamapattam, only very little can be inferred. That he
belonged to the Cetti class, is evident from the class appellation found
in the copper plate to follow his personal name (cf. Ifavikorttaiianaya
f'eramanlokappefuncettikku). From the grant it is also clear that the
donee is a Makkattavi, i.e. one whose succession goes to his own son.
as opposed to Mafumakkattayi ” one whose succession goes to his
958
K. OOOAVARMA —
sister's son As Ccttis also are Makkattayis. Mr. Daniel s iiit n
that Ifa\’ikorrtan must be a Christian on the ground that ■ l^
referred to as a Makkattayi. cannot stand. In conclusion it i "p
said that the donee was a Hindu by name Kavigovardhana. ar. aar
he was a member of the Cetti class and a resident of ]\Iak'- car
Pattanam.^
Ill
Grant
The grant made by Vlrafaghavacakravartin to Ravigovar ; ^na
is Manikkifamapattam. As to the meaning of the term Manikkii a^
pattam, opinions of .scholars differ. Dr. Gundert observes : "AS ’mii
tradition relates that the Syrian Christians sorely tried by a lu a 'ii
conjurer (the poet IManikkavacakar ?) at last divided into two pa '
one of ninety-six families, who submitted to heathen puriticatioi. iid
adopted the conjurer's tenets, whence they were called Manigfanial n
etc. " (Madras Journal, vi. xiii, part i, p. 121). Later Dr. Cui '
remarks in a letter he wrote to the Reverend Taylor, comiut t ' laa
upon the views expressed by the Reverend Peet, of Jlavelikaray a-
follow.s : “I now venture the conjecture, that they were discipl'-
Mani, a colony of Manicheans. such as the Arabian Travellers ii>’i '
in Ceylon " (Madras Journal, vol. xiii. part i. p. 146). Kookel K-
Xair explains ilanikkifama as the village called Mani, the n<i'‘
according to him, being derived from the wealth it then pos.se'>>
Mani, he says, may also mean best or chief (Madras Journal, vol. ■
pp. .35 and 41). ilr. Daniel interprets Manikkifamapattam as the tr.
of Manigfaniakkafan. which, he .says, might mean a valued eiN .
privilege of the time. As regards the etvmologv of the word. 5Ir. Dam
is silent.
We shall now examine the above vi('ws one by one. Concerning tie
opinion that IMani is the name in memory of ^Miinikkavacakar, it lae
to be remarked that the tradition is only that the Christians relaps' i!
into heathenism through the influence of a sorcerer. The inckh'ii’
^ Miikotaiyar Pattanam, says Ur. Giindert, " is klentiHed with Tihivaiicikkula"
river harbour, lat. 10 13' N." The place is also known by the names of Maha<le^Al
Pattanam and Mahodevar Pattanam. Makotaiyar Pattanam is obviously from
Mahodaya. Mr. Daniel's interpretation of Makotaiyar Pattanam as the town of Ikt
Great Cera King cannot be justiHed, inasmuch as a combination of " maha ’ an'!
‘ kotai would give rise to a form " Makotai " and not Makotai Mahade%ai
Pattanam is probably a name given in reverence of the presiding deity of the plai^ ■
while Mahodevar Pattanam is evidently a name that has been formed by cross analog}
of Mahodayar Pattanam and Mahadevar Paftaiiam.
COPPER PLATE GRANT OF SRIvIRARAGHAVA CAKRAVARTIN 959
was associated with the name of Manikkavacakar by Dr. Gundert and
the Reverend Feet mainly on the external resemblance of the two
words Manigfamam and Manikkavacakar. This is evidenced by the
interrogation mark Dr, Gundert has used after the word. The change
of Manikka{vacakar) to Mani. even if it be contended that it is an
abbreviated form, is against the principles of all sound changes.
The shortening of the long vowel of the initial syllable and the elision
of the final letter which forms, so to speak, a prominent member of
the word are against the phonetic tendencies of ilalayalam. The
inference Manikkifama took its name from ^lanicheans pre-
supposes that Manichean missionaries visited Kerala. Such a
presumption is unsupported bv any historical evidence. Further,
the ancient Christians of iMalabar are found from history to have
been followers of true Christian faith unlike the ^lanicheans of mixed
religious beliefs (see Indian Antiquary, vol. iv, p. 311). The meaning
that Kelu Nair gives is out of the question, being far-fetched and
impossible to derive : Manigraina can never connote in Sanskrit the
sense that he makes out. Nor can Manikkifamapattam be a mere
civic right as suggested by Mr. Daniel, inasmuch as the rights and
powers conferred on the donee are of such a supreme order as are not
likely to be bestowed upon every citizen of the place.
Let us now consider what sound changes may have legitimately
occurred in the word. That initial v and 6 in the Dravidian languages
may sometimes be pronounced as m when there is a nasal in the body
of a word, is illustrated in the following examples : —
Earlier Form.^ Later Form.
(Tam.) mannan (Mai.)
hanUuvu (cf. H.M. bdyhi) inahkiavu (Mai.)
nljinyu (Tam.) mihiiinn (Mai.)
>'dnam (Tam.) inCtmun (Mai.)
>'im (Tam.) miha 'm mimkkPtu (Sled., ref. Gundert)’
It is thus clear that Manikkiiama must be derived from Sk.
Vaniggrdma,^ which means an as.sociation or guild of merchant.s.
Attention may also be drawn to the meaning kahn'in per (name of an
association or companv) given to tlie word Vanikkifdniani- by
Nacciiiarkiniyar in his commentary on Tolkdppiyam. The fact that
^ That the forms quoted represent an earlier &tage in the history of the words in
question, can be known through comparison.
- In Vaniggrama, griima means only a collection (cf. suffix grnmnc of Panini) and
not a Tillage as interpreted by some.
960
K. GODAVARMA
the name Manikkifamakkar denotes a certain section of t:, Xair
community as well, goes against the possibility of its l»i j an
appellation of the Christians alone. The rights and privilege' 'rail
upon Ifavikorttanan will be seen to be those connected with liTu
and inland trade, the lev’ydng of duties and kindred matters pen ,iii»
to commercial transactions. Manikkifama is also found to hav -en
used as a name of the Indian colonists in Siam (ref. the artn A
note on Manigrama " in Epigraphica Ivdica. vol. x\'iii. j '••)
Manikkifama. therefore, does not admit of being interpreted - ;li''
name of any particular village or community. What Ravigovai. oui
received must have been only the lordship of the trading corpei
which transacted business at Kotuhkulur during the da\' "f
Vlfafaghavacakravartin.
IV
Interpretation
In this section I propose to give my interpretations of c ' in
expressions worthy of note in the document. I have shown iUi’ I't
each word the views of others who have worked in the line, and ■ v'
also tried to accommodate myconclusionswith the phonological cli.i -
manifested in each case.
(1) Vila I’d ta : Gund. "feast cloth"; Kel. pildvdfa = " n ' ' ‘
permitted to be worn round the shoulders by people of high ' a-
alone ; Venk. “ the right of festive clothing " : Dan. " the riglif
the feast-cloth " (derived from the word ri/n meaning " festival
the chief gue.st during the marriage festival of the St. Thou- a
Christians is seated on what is called veUaymn' knlimpatavinii).
There is no word vildvu in Tamil. Besides, it has to be inferiv
on the strength of the evidences offered by other words in the copi>'
plate, that the change / to I was not in vogue at the time in Malayahm
Cf. vjydlat/i. aliviyotu, ehili, ralimUye. Even to-day Malayalam. <1
all the Dravidian languages, has pre.served / without confusing it witi.
L If the original form of the word is vildvdtn. meaning cloth covcrin-
the sides of the body" (rildru =“ sides of the body"), the vow i
in between r and I could elide as in ktdvii < kitdvii, ptdka < pitdkii-
lidkam < rifdkam, etc., and the I in conjunction with a consonant
could easily change to Z as in pldsu (l.w. from Sk. paldsa). pldvu (earln i
pildvu). The consonant group may, at a later time, split itself int"
two syllables with the development of a vowel in between. Ih<'
COPPER PLATE GRANT OF SRIVIRARAGHAVA CAKRAVARTIN 961
existence of words like pildcu, pildvu are instances to illustrate the
point. When we remember that the wearing of upper garments on
public occasions was a privilege bestowed only upon persons of high
rank and dignity, by monarchs in ancient Kerala, the meaning that
1 have suggested will be found to suit the context.
(2) Paranattdnku : Gund. ” house pillars ” or “ pictured rooms " ? ;
Kel. "a carved pillar”; Venk. “house pillars” ; Dan. “house
pillars ” ; Jos. means perhaps “ an enclosure around houses Shoidd
the reading pavanittdnku be accepted, the word would mean protection
for pavani, i.e. praydna.
The meaning “house pillars” or “carved pillars” is only con-
jectural. besides being uusuited to the context. The erecting of pillars
for houses cannot be, in any way, considered a mark of s.uch a great
dignity as to require royal sanction. Mr. Joseph’s interpretation is for
the reading Pavanittdiiku, which he has adopted. I am inclined to
think that Pavanattdhku means support for the house, i.e. sentries to
keep watch over the house. To a person who has been raised to such
a responsible position as that of the lord of Vaniggrama, in which
capacity he has to keep large amounts of money in his house, it is
only proper and reasonable that the necessary safeguards are allowed.
(3) Peruperu ; Gund. veruperu “ all the revenue ” ; Kel. pufupcr
= ■■ all the revenue ” ; Venk. pent per — “ income that accrues ” ;
Dan. “ all the revenues ” ; Jos. veruperu “ other privileges ”.
The readings of Dr. Gundert and Mr. T. K. Joseph are due to a
mistaking of v for p. Kelu Nair seems to have purposely read the word
as pufuper to inject into it the idea which he had in view. I agree with
the reading and meaning given by Venkayya and Mr. Daniel, peru,
which appears as the initial member of a compound, means “ that
which accrues ” and peru means “ income ”, the whole compound
conveying the sense “ all the revenues
(4) Kaluttuvalenciyam : Gund. “ the curved sword ” (or dagger) ;
Kel. “ ferry tolls ” ; Venk. “ export trade” ; Dan. valenciyam may
mean something akin to feudal barons suggested by Simdaram Pillai ;
Jos. " export trade ”.
Only Mr. Venkayya has made an attempt to trace the derivation
of the word. He draws attention to the various forms of the term
ealanciyar appearing in Ceylonese and Mysore inscriptions, and quotes
related words from Kanarese and Telugu. The words cited by Venkayy^a
are vlrabalanji, bananjika, hananju dhanna (Mysore inscriptions of
^Ir. Rice, Nos. 38 and 55), Kan. banajiga, and Tel. balija and balijiga
962
K. UOOAVARMA —
(a class of merchants). These, he thinks, are cognate and are ii- ;
from Sanskrit vanij-.
I fully accept the meaning given by Venkayya. As regar '' ‘i -
etymology, I would connect the word with Sk. vanijija and accuir. i
its present form by assuming the following steps in the cour'c ■
evolution.
Sk. vanijyd, Pk. rayijjd and *i'anijjiht (with the exteii'i' ' e
ika of the three suffixes of extension in Prirkrit. namely, aht. '/,e. ;
uka). The Prakrit form must have been borrowed into the iMal.i\ '
vocabulary through Kanarese, for the change of -n- to -1- is a te
of Kanarese alone. Kan. *vauik>jika > banijika (probably intlui :
by a new Indian dialect where the simplification of long congou ■ ■■
with the nasalization of the preceding vowel is very common (ct
Knbja-, Pk. Kiijja-. B. Ab/j ; Sk. yndhya-, Pk. jnjjha-. ,M.
Kan. *banejiya {i when followed by an open vowel in the next syP
regularly changes to e in the colloquial dialects of the Dravio .
languages, and -y- frequently elides). Kan. *bal?jiya or bab i'^
(for the change of -)i- to -I- in Kanare.se see Kan. anil, afil " scpiiii'
>Ial. anndn ; Kan. itni, nli " one who eats ", Mai. an- ; Kan. tmnn
tulaku ‘’fragment", Mai. tnni). Mai. valenciyam ot ralanciyatn i' '
the substitution of r for 6 of a foreign language in Malavalam see M
villa for H. billd ; Mai. vllam for H. bldkd ; Mai. rattam tor H ■
batta, and Mai. vdnnu for H.M. bdk).
(5) Valanciyattil taniccettam : Guiul. ” in the sword the .soverei.
merchantship ’ ’ ; Kel. “.sovereignty or executive jurisdiction
Venk. Dan. "monopoly of trade"; .Jos. "sole monopoly '•
trade ".
As the meaning of the term ralanciyant has already been discus' ll
we need consider here only taiiiccettu. The word is used in the seii'C "i
sole monopoly of trarle. Celia goes back to Pk. .'^utta-, derived from Sk
sdrtha- “ a travelling company of traders " (cf. Mai. can iKilnii' .
cavalam. and canhala, borrowed from Pa. sahyhdta-, Pk. savcahn
and sahkJiald respectively, exhibiting the change of ludo-Aryan to
c- ; e for a after c is a colloquialism in Malayajam which is evidentl}
due to the palatal position of the tongue required for the production ot
the consonant influencing the vowel that follows. See Sk. campnka -
Mai. cctnpakaiH : ilal. carivii, cefivit ; ilal. call and cell ; cf. Kan.
soli ; Pk. camma-, iMal. cemmdn ; Pk. cakka-. Mai. cakku or cekkn)-
For the expansion and contraction in meaning of the original Sk.
sartha- in the course of its hi-storv in different languages may be
COPPER PLATE GRANT OF SRIViRARAGHAVA CAKRAVARTIN 963
compared H, sdtd “exchange”, P. sattJ "exchange market”, G.
sdtu “bargain’’, etc.
(6) Pdvdta : Gund. " the spreading cloth " ; Venk. “ clothes spread
in front to walk ” ; Dan. “ clothes over which lamps are placed ” ;
Jos. by pdvdta is meant natapdvdta or “ clothes spread to walk ".
The word pdvdta is not Dravidian in origin. It is derived from Pk.
pdvada- traceable to Sk. prdvrta-. The long vowel after v in Mai.
])dvdta is due to a frequent tendency in Malayalam to lengthen the
vowel preceding the last syllable.
Cf. Sk. kdnhold
Pk. pattaya-
Pk. bandhaya-
Pk. riakka-
Mal. kdmdla
Mai. pattdyani
Mai. panfdyam
Mai. vtydkku
Dr. Gundert gives the meaning ” table cloth or other sheet used
to fan " to pdvdta. A similar development of meaning is to be seen in
Kanarese (see Kittel’s Dictionary : pdvada = " a cloth waved like a fan
in front of a procession "). The word, by reason of its being enumerated
in the text along with the honours and privileges the lord of vaniyyrdiaa
was to enjoy in his processions, is better interpreted as a sheet used to
fan or a cloth waved like a fan in front of a procession.
(7) Itupvtiitoranaw : I take this as one word instead of two. as
explained by others. The reason for my so doing, is the absence of the
connective am after itupati as in rildvdteyum, pavanattdiikam. eJc.
Itnpatitdfamm means festoons over gateways". The lord of
vnniygrdma was to be honoured by people during his processions by
ornamental arches or festoons over the gateways of their houses.
V
Language
The language of the copper plate is okl Malayalam. There are to
be seen in the text a fairlv large number of Sanskrit and Prakrit loans,
some as Tatsamas and others as Tadbhavas. while New Indian words
are totally absent. From a .study of the Sanskrit and Prakrit loans
we can easily discern that the intial media and media aspirates are
represented as tenues in the copper plate, ex. kirdmam for Sk. (inlma-,
jKivanam for Sk. hhavaaa-. and pupdla for Sk. bhdpdkt-. The same in
the intervocalic position appear in certain instances without the above-
mentioned change. See vlhudyhavan. caadmdityakaf (without cluinge).
and dtiydyi. makdtaiyar. imrikdritamni (with the sulistitution of tenues
(12
V(1L. vm. PART 4.
964
K. GOBAVARMA—
for media aspirates). From this it could more or less be infen. ■ nt
against the then general tendency in Malayalam to pronounce ■■d
sounds as voiceless, the sounds to be pronounced first as voie>' e,
true to the original, may have been those which occurred in b. r . n
vowels, in w hich position ev’en voiceless stops are pronounced \ J.
(The v oicing of intervocalic stops is a regular phonetic ph enomei- in
Prakrit and Dravidian.)
Another interesting phonetic peculiarity that we may nob. is
the confusion between c and s, original c sometimes being repres. .1
as s and vice versa, ex. sakravartti for Sk. cakravartin- (s for c). cu; .r
k. sa«i (c for s). It may reasonably be conjectured that i of San : t
was at one time represented as c in Malayalam and that later t u-
arose in Malayakm a tendency to pronounce c- as s~ colloquially. >'■
modern Malajajam colloquial words sTl/u for cittu, sappan for capj
sappaUa for cappatta, and satni for eatni. The language of the plate , . '
points to the fact that Prakrit had great influence over Malayalam at •
period in question. F or want of comparative study, the Prakrit eleni. r
in 3 Ialaya!am has not received careful attention of the scholars i ■
have worked in the line previously. I give below the Prakrit borrowii . ^
m the copper plate with their Sanskrit equivalents •—
Mai. form in the Plate. Prakrit. Sanskrit.
seUhi- sresthin-
paUinam patlana- pattana-
saUa- sartha-
pavada- prdvrta-
vdnujar^ vdniya- vdnija-
kamtndar *kammdla- karmdra-
cunkam suhka- kulka-
e above words cannot be considered as borrowdngs from Sanskrit,
since t ley are not governed by the rules regulating the change ot
1 adduced by philologists. For example,
1 T * ® ■'should naturally give a form cifetti in Malayalam
and laiml if it were a direct borrowing from Sanskrit. The absence
nmv ndian words goes to prove that at the time of the document
the New Indian languages have not had any tangible influence over
Malayalam, although they had by that time established themselves as
independent dialects.
Regarding
assimilation or
the Dravidian words, it has to be observed that nasal
o-hundsikdtiprasafa, w'hich is a characteristic feature of
COPPEE PLATE GRANT Ol’ SEIVIRARAGHAVA CAKRAVARTIN 965
the Malayalain language, had not manifested itself in all combinations
at the period in question. It is only nt that shows assimilation while
nt and hk remain without change, ht and nt of ancient Malayalam
have, however, fallen together and have become nn in modern
Malayalam. We can also infer that the first combination to take the
assimilation was nt and the second in order was iik. There is every
probability that the copper plate belongs to a period prior to that of
I nnunilisandesam, in so far as hk as a whole is met with in that work
only in its assimilated condition, that is hh. Cf. tahinttahin, eniial,
nuruhhu, etc., in UnnunilTsandesam.
VI
Date
The only material in the copper plate which affords some clue to
its probable date is the astronomical positions mentioned therein.
It is found that the document was executed on a Saturday, which
was also a day of the fourth lunar mansion, that is Rohani. and on a
twenty-first Miham when Jupiter was in Makafam. Kookel Kelu Nair ^
ilated the plate 6th March, a.d. 230, as he found, on calculation, that it
suited the astronomical requirements referred to. Dr. Burnell,^
after consulting native astronomers, arrived at a.d. 771 as the date
of the grant. Keilhorn ® corrected the above date to 11th March,
A.D. 775, to be precise, and also suggested that 10th March, a.d. 680,
would also fulfil the astronomical requirements. Venkayya, on
palseographic and linguistic evidences, has tried to prove that the
document must have belonged to the fourteenth century a.d.
{Epigraphica Indica, vol. iv, pp. 292 and 293). Later, Keilhorn, in the
light of Yenkayya's evidences, pointed out 15th IMarch.* 1320, as the
only date in the fourteenth century which will .satisfy the astronomical
clues. In recent years Mr. K. N. Daniel is seen to have taken
considerable pains to ascribe the document to a.d. 230. On an
examination of the language and script employed in the copper plate,
I am inclined to think a.d. 1320 to be the date of the grant. In this
connection I should like to refute tlie main arguments brought forward
1'}‘ Mr. K. N. Daniel in support of his theories and against the views of
\ enkayva.
^ ^ladras Journal of Literature and Science, vol. xxi, p. 39.
- Indian Antiquary, vol. i, p. 229.
^ Ibid., vol. xxii, p. 139.
* Epigraphica Indica, vol. iv, p. 293.
966
K. GODAVARMA —
It cannot be argued from the fact that the document undi !!-
sideration uses the form kututtom nine times, but nev^er kotuttdn ' at
it is a very ancient one. Mr. Daniel has proceeded on the assiir >11
that kutu is the earlier form, and kotu, the later. I should thi.n at
kutii is only a colloquial pronunciation, and as such it is not helj m
any way to determine the comparative anteriority or posterior ot
the plate. The change of m to o occurs when there is an open ‘ ' <'1
in the succeeding syllable, ex. urahhuka, oraiinuka : uta
oruifuka ; kuta, kola ; lufakkttka, tofakkuka ; mutantart, nioi’ '•
On the other hand, when there is a close vowel in the syllabi. at
follows, no such change is possible. See Kulikkuka, tuni. uonl "
ufuli, utuppu. It may be remarked that there is a probabilit\
being made closer when there is « in the nc.xt syllable. Henc "
may show a form kutu in the colloquial dialect. Kututtom. then
can only be a colloquialism that has crept into the copper plat.
Mr. Daniel quotes two south Travancore inscription.^ ol
thirteenth century, a.d. 1237 and 1251 (Dissertations on the
Plates in possession of the St. Thomas Christians, p. 21). and comi’ ■'
the specimens of the Malayajam prose therein with the language ot
Vifafaghava plate, and obser\'es that we cannot help concludin'' f
the copper plate is many centuries prior to the thirteenth cent'
But a careful comparison of the inscriptions referred to with
Virafaghava plate, will show that there are more points of simila
than difference. The following are a few words that are found tn
common to the Vifafaghava plate and the inscriptions un.
reference : viydlam, hdyaru, celld (yi) ninta, epperpetta.
The plate, showing the variations of the Arya elutlu alphab.
given by Mr. Daniel in his book, is in no way calculated to show that tl
characters in the Vifafaghava plate are more archaic than those foum
in the Tanufavi plate. It may al.so be observed that the Arya charact. i
employed in the Vifafaghava plate resemble best those used in th'
inscription of Rajendra Coja, dated A.n. 1012. Above all, a mere glam .
at the table will show that, of the sixteen types illustrated in the plate
the one that is nearest to the modern Malayajam alphabet, is that o'
the Vifafaghava plate.
Mr. Daniel’s contention that Vifafaghava is earlier than
Parkafafavi, on the ground that Vifafaghava is a pure Sanskrit name
while Parkafafavi is Tamilized Sanskrit, can in no way be acceptable
to those who are familiar with the history of the influence of Sanskrit
^ a is an open vowel while i and u are close vow els.
COPPER PLATE GRAN’T OF SRIVIRABAGHAVA CAKRAVARTIN 967
over ilalayalani and the mutual relationship of Tamil and Malayajam.
Of the two names Parkafafavi and Virafaghava, the one that is
evidently older is the former. This is proved by the representation of
voiced aspirates without change in the name Virafaghava and the
substitution of tenues for the aspirate in Parkafafavi.
Palseographic evidence, Mr. Daniel contends, especially that afforded
by the character lu of vattelutlii, also goes to prove that Vlfafaghava
is earlier. From the plate showing the variations in the vattehittu
alphabet given in Mr. Daniel's book it seems that Mr. Daniel
is labouring under a misapprehension. Otherwise he would not have
attempted to draw a comparison between the ]u of the Vlfafaghava
plate and that in the Tanufavi plate, etc. It is not the hi found in
Vlfafaghavacakravartin's plate that has developed into the various
types illustrated in the table given by Mr. Daniel. The truth is that
the lu in numbers 1, 4. 5. and 21 is valtfluttii. while the remaining are
variants of a different kind altogether. They are in reality different
forms of Devanagari I (?«5) with the addition of the symbol for
cattehittu IL. That Devanagari has been borrowed by other
Dravidian languages also, is evidenced by the script in Kanarese
to denote consonant group as M. yh etc.
I. therefore, think that of the four dates satisfying the astronomical
recjuirements, a.d. 1320 is the probable date of the grant, in the light
of the linguistic and palseographic evidences available.
COXCLVSION
To sum up, the purport of the document is the grant of the lordship
of Vaniggrania with the rights and honours pertaining thereto to one
llavigovardhana, a member of the Cetti community and a resident
of Mahodayarpattanam by Sri Vlraraghavacakravartin, on Saturday,
the 15th March, a.d. 1320.
A propos du genitif absolu en vieil indien
Par H. DE Willman-Grabowska ^
T E probleme de cette espece de genitif a ete traite a fond par
F. de Saussure (De I’emploi du genitif absolu en sanserif, 1881,
reimprime dans le Recueil des publications scientifiejues de F. de 8.,
Heidelberg 1922). Au debut du memoire on lit : Un premier fait,
constate depuis longtemps, e'est I'absence du genitif absolu dans les
monuments de la periode vedique. ’
La notion vedique ” ne comprend pas uniquement. chez F. de 8.,
la periode des Saiiihitas, niais s’etend plus loin, sur les Upanisads. Aussi
1 auteur tient-il compte de la Maitrayanlya-Up. et analyse le passage
ou intervient misato bandhuvargasya, cite par le Diction, de Petersb.,
mais refuse a cet exemple la validite de temoignage. Remarquons
cependant c[ue la Maitr.-Up., texte tardif. proche par sa langue du
Sanscrit classique (cf. H. Oldenberg, Zur Geschiclite der altindischeu
Prosa, p. 33), ne saurait rien prouver pour la periode ancienne.
L’absence ou la presence du genitif absolu est Taffairc du style, propre
a ce texte.
Delbriick (Altind. Synt., p. 388, s.) est moins affirmatif. D'accord
avec Gaedicke (Der Accusativ im Veda, p. 47 ; Breslau. 1880) il
voit un genitif absolu dans la phrase du SB. 1, 1. 4, 15 : tdsySlabdhasya
sa vdg (ipacakrdina ^ et traduit le groupe au genitif non pas comnie le
complement d'objet indirect (le gen.-possessif). mais conune le com-
plement circonstanciel ou plutot comme une proposition circon-
■stancielle subordonnee : " apres qu’il fut immole.’’ Du ineiiie avis
est J . Eggeling : il rend ce passage par ; " on his being killed the voice
went from him.” Aux exempies releves incidemment par Gaedicke,
Delbriick ajoute plusieurs autres. Ils se lais.sent tons analyser de fa 9 on
lormelle, grammaticalement, comme des complements de nom, mais
tous voisinent en meme temps avec I’emploi du genitif absolu tel que
1 indique de Saussure pour ses groupes A et B (Recueil, p. 278 s.).
Ce n’est cpie le voisinage, car la notion de genitif abs. de Altind.
bynt, est un pen differente de celle que pose " De I’emploi du gen.
abs. ’ De Saussure, a rexamen de I'epopee et de.s textes classiques,
arrive a la constatation ; ■' L'action du geuitif absolu accompagne dans
' This article was intended for Vnl. VIII, 2 .iiui 3, the volume dedicated to
•'■ir George Grierson, but it unfortunately arrived too late.
Cf. la phrase tout a fait analogue de XaJa tt Damayanti, xiv, 12 : la‘‘ya d<iMa-->ya
hid rupaiii ksipram antaradhiyata.
970
H. DE WILLMAX-GEABOWSKA —
le ttmps I'aclion principale ^ ; la premiere n'est jamais doiiiRt ' im
close au moment oil la seconde s'accomplit.’’ Le sens d apre.^ '
exclu. Tel n'est pas le cas dans les tournures du type de tfhijnhil- -pi
. . . on de apSm taptdnam pheno jayate (A.S., p. 390). Non m u • ut
Taction du genitif prec^e ici Taction exprimee par le verhimi " nin
(a pacakrdma, jayate), mais elle doit etre terminee avant ipic la da
commence. Nous somme.s ain.si dans le domaine occiipe ave* m
succes par le locatif absolii.
De Saussure constate dans le genitif absolu de son groupc A 'a
sorte d’arriere-plan sur lequel le fait principal se detaclie. 1 • '1'
groupe B, inverse de A, le genitif renferrae " une circonstancc cs'i'i Ha
a Taction et cette derniere souvent, mais pas toiijours — .-e "a
nialgre Texi.stence d une entrave du fait de cette meme circoio '•
essentielle. Ce dernier cas, est seul mentionne par Panini d i !•'
sutra ii, 3, 38. mstfn canadare.
Ainsi d'un cote Tapplication restreinte et tout a fait partu '
speciale, du genitif absolu. designee par le granunairien indii" da
Tautre la notion clargie, due a Texamen des textes postcriai.
Panini, les epopees, les Puranas et les contes. creations vivin'.t' - i*
partie populaires ; car c'est la surtout que de Saussure a tai'
recherches. L'observ'ation penetrante de Gacdicke (op. fit.) '■ H
Delbriick (op. cit.) vint ensuite enrichir notre connaissance du mi'
Cependant a niesure qu'on elargit la notion de genitif absolu.
parait moins nette et on ne saurait toujours distinguer oii conmia
le genitif indepcndant et ou finit le genitif possessif on autre. i
contexte n'est pas dans tons ccs cas une garantie suffisante. car n
ne sommes C[ue trop enclins a juger de la relation entre les grouy
d'une phrase d'apres nos habitudes de penser et de parlor (" • •
view the syntactical relations of one language through the disturlun,
•medium of another " — H. Oertel, The Syntax of Cases in the narratr
and descriptive Pro.se of the Brahmanas).
H. Oertel dans son etude minutieuse des Brahmanas et '
Upani.sads que de Saussure n'av. pas analyses, cite beaucoup d exenipH
du genitif qui pourrait passer a la rigueur pour absolu sans leto
certainement. II s'arrete en fin de compte sur un petit nonibre
quelques uns deja releves par Gaedicke et par Delbriick — pour H"
classer commc de vrais genitifs absolus.^ Ils ne disent pas toii^
“ pendant que ” ; plus souvent ; “ apres que ” ; aucunnedit; “ lucu
que ; ils sont done contraires a la regie de Saussure et a cellc di
* Souligne dans le texte. 2 p. 142 s.s., op. cit.
A PBOPOS DU GEXITIF ABSOLU EX VIEIL IXDIEX
971
1‘aniiii. Et neanmoins I'auteur a raison de ne pas voir en eiix qne
des genitifs possessifs.
On est cependant etonne de ne pas trouver dans cette petite liste
deux exemples, on plutot un en double, du Brhad-Aranyaka 2, 4. 4,
et 4, 5, 5. Le groupe au genitif designe la une circonstance eon-
teinporaine a Taction, nominee — injustement mais habitiielleinent -
■■ prmcipale ’’ (le groupe A chez de Saussure). il la met en relief et
re^te en meme temps condition essentielle pour qu elle puisse se
protluire (le groupe B) ; de plus, le sujet du genitif est le nom d une
])er.sonne. Autant de traits valables pour un genitif absolu.
Maitreyi demande a Yajn. de Tinstruire. Y. repond : e/i// iisra
i ijakhyasyawi te vyacdh^anasya hi me n'ldidhydsasva Br.Ar. 2. 4, 4.
■‘viens. assieds-toi, je vais t expliquer, mais pendant que je parlerai.
'<iiis-moi bien avec ta pensee L'action du verbe dhyai ne pent se
derouler qu a condition de voir s'exercer eelle du verbe vi-a-caks. elles
sont simultanees mais il faut que Tune commence avant 1 autre ;
avant que Yajn. ne se mette a parler. Maitreyl n’aura pas a rMechir.
et Taction " principale " est commandee par Taction secondaire ,
elle lui est subordonnee. Il est evident que cette subordination est
le resultat de la pensee entiere du passage : elle n‘a rien a voir avec
le genitif meme. — Le fait est analogue a Tindependance de 1 anadara
du genitif C|ue constate de Saussure p. 280. — On ne saurait considerer
eyucah^dnasya vie comme Tobjet du verbe dhyai ; 1 objet serait me
(si 1 on veut construire dhyai avec le genitif). et ryacah-fa nasya est
1 attribut de me. Mais la notion qu'iinplique ce genitif absolu de Br.
Ar. permet un emploi beaucoup plus large qu on ne le constate plus
tard. dans la litterature classique.
L'autre exemple c’est le meme passage elargi au moyen d un sub-
stantif Br. Ar. 4, 5 ; idcam hi me fydcdkmnasya nididhydsasi'a. La
presence de Tobjet direct vdcam. olijet du genitif vyScaksanasya et
du predicat nididhydsasva ne change rien a la valeur de la
tournure meme.
Oes deux exemples oil Taction du predicat est simultanee a cede
du genitif, sont a peu pres exceptionnelles dans la vieille langue. On
les a construits aussi regulierement que iti cintayatas tasya ayayuh
siriyah Kath. xviii. 356. ou na hi team jivato tasya eanam- ayantum
arhasi Ram. ii. 101, 3. cites par de Saussure (op. cit.) et par J. S.
Speijer (Sanskr. Syntax, p. 228).
La regie generale C|ue le gen. abs. est applicjue avec le sens premier
de '■ au moment oil ” n'est pas toujours observee non plus, meme
972
H. BE WILLMAX-GKABOWSKA —
dans la litterature epique qui fournit le plus grand nonilirt' dc i a-;.
Ainsi dans le P chant d'Adiparvan, vers 7 du 3IBh., chant .1 aine
incontestablement tardive, les rsi voudraient entendre ran. r le
MBh. , 1 un d eiix interpelle longuement Suta et finit par ei ' jTs
samsaitat prcchato mama " Dis-le (on : recite-lel puisque jt 1'‘
demande ou: '■ du moment que je te le demande comm an
dirions simplement : =• je fen prie." La relation entre le gei . ft
le \erbe nest ni pendant que ' ni “ quoique c est '■pui'"'.
prcchato . . . est la subordonnee de cause. Les sages de la furet A.
■sont superieurs a Suta. cf. les slokas 8-13. II n'oserait leiir i)aii i
ne 1 j autorisaient pas ; prcchato . . est une forme polie de comm.
ment. On trouverait sans doute plus d'exemples de cetti- l;'*
application du gen. abs.
Ainsi la prose post-vedique et la jKiesie epique perniettei f
conclure qu il existait des tournures composees d’un participe pr
et d un nom, tous les deux au g^nitif, et ayant la valeur d'uii ■
plement circonstanciel ou d une circonstancielle subordonnee de t' . [<-■
ou de cause, tournures qu on doit compter parmi les genitifs ali'' i'
Maintenant si 1 on se tourne aux Vedas, on y rencontre des pin
qui ne sont pas sans interet pour le probleme qui nous occupe.
Ainsi R\ . viii, 37, 7a Syavah'asya sunvatds tdtha sryu ydthtu ■
dtreh Mrmam Icrnvatdh peut etre traduit : “ ecoute S. pressuran'
Soma (qui pressure ? quand il pre.ssure ?) comme tu ecoutai.s. '•
quand il offrait le sacrifice," mais le contextc et d'autres pas.sage^
rapportant a Atri ne nous interdi.sent point une traduction coim
celle-ci ; ecoute quand S. pressure . . . comme tu eeoutais lorsi;
meme viii, 35, 19a dtrer iva srnutaih purvydstidi •
syaidhasya sunvato madacyutd “ I ous deux, o ruisselants de mail,'
ecoutez 1 ancien hymne de S. quand il pressure le Soma, comme voii'
avez ecoute . . . etc."
Quand il . . . s impose ici pour traduire siunato. Quant au
que srx pent avoir fobjet ou genitif (ce .serait alors un genitif partitif )
fait note par les dictionnaires. il ne change rien. Autant que j ai<
pu voir, e'est a cote (run participe pres, (qui n'indique pas imce.ssaiiv
ment le present mai.s .seulement la simultaneite) que h-n a et (3 relev.'
construit a\ec un genitif; dans tous les cas verifiables ee genitif dn
partmipe pouvait etre traduit " pendant que ".
D autres exemples sont plus douteux ; ainsi RV, 38. 9h, gopdjih-
feT vih-c pasynnti mayinah krtdm "de G.. quand il
’ dieux {mayinah) voient les diver.ses creations ’ est
A PROPOS DU GEXITIF ABSOLU EX VIEIL IXDIEX
973
incertain, car le contexte, qni semble avoir pen de rapport avec cette
phrase, ne I'eclarcit pas. Du reste le participe present est a.ssez
peu employe dans le BV. ; ses formes sont nombreuses, leur application
est limitee ; le participe au genitif est le inoins frecpient, et la oil
on le rencontre il apparait le plus souveiit en fonction d'adjectif.
Ainsi la matiere a construire le genitif absolu fait dtfaut.
Le butin de TAtharva-Veda n'est pas riche, non plus. A\'. iv, 2, 8.
parait sur ; apo vatsdm janchjanfir (j(irbha)n dgre samairaijan — tdsijohi
jdijatndnasyolba asld dJiiranyuyah ; Whitney et Lanman (HOB. vii)
le traduisent The waters, generating a young, set in motion in the
beginning an embryo ; and of that, when horn, the foetal envelop
was of gold Aussi AV. xix, 62, 1 : priydm md bnni devesn priynth
nija.ni md l-piu 1 priydm sdrvasya pdsyala utd siidrd nlGrye " rends moi
agreable aux dieux, . . . aux rois . . . aussi au sudra et a 1 arya, en
presence de tout (ce Cjui vit).’" Je me permets ici de ne pas etre
d'accord avec les traducteurs qui disent ; dear to everjdhing that
sees, both to Sudra and to Arya." La strophe entiere construit priydm
Did Irnu avec le locatif (devepf. rdjasu, siidre, iirye) : il y est repete
quatre fois a titre de precede stylistique. L'auteur pourrait tres bien
conserver la mesure du vers avec sarvasmin . . . s'il desirait dire " a
1‘egard de, etc." ; I’emploi du genitif. seul parmi tous ces locatifs,
etait provoque par des raisons du sens : je veux etre agreable ... a
la face de tout," textuellement " quand tout (me) regarde ‘ . C est
un genitif absolu.
On en trouverait plus d'exemples, mais I'extreme concision des
tommies et le stvle poetique du BY. et de I’AY. accumulent les functions
qui devraient appartenir a plusieurs mots et en chargent un seul. C est
-surtout frappant pour le genitif : il lui arrive d etre a la fois gen.
possessif, gen.-datif, gen.-ablatif et gen. absolu. tout cela selon la
maniere dont on envisage I'ensemble.
Bi peu qu'on puisse tirer des Yeda.s, leur examen est instructif.
Lorscju on est en presence des phrases-types (cette espece etant tres
repandue) comme suivantes : sHab sdulo hidayani f ir dahanli x 34. 9h,
" (tout en) etant froids, ils brulent le coeur,' oil le participe est
1 attribut du sujet, et jciyd Uipyatr kitardsya iiiva mdta putrasya
edratah I vd svit x, 34, 10a : " la femme et la mere du joueur se de.solent
abandonnees. pendant qu'il erre on ne salt oil. oil le participe attribut
d objet peut correspondre a notre phra.se circonstancielle,— on assiste
a la naissance du genitif absolu. La formation des tournures plus ou
moins autonomes. equivalences des propositions subordonnees, etait
974
A PROPOS DU GEXITIF ABSOLU EN VIEIL IXDIEX
ine\itable en vieil indien. La langue des hymnes, fie bon: - 'iMurH
stabilisee, etait le seul modele de parler noble. Mai.s c'etait ],i
de poesie. Son but n etant pas de faire comprenclre mais tie fau' ntir
de\iner et voir, enfin d’eveiller des associations : elle evitait it tf
qui pouvait enlever a la phrase son caractere rapide et aile. Cej lint
pour tracer en grandes bgnes, et de fafon concise. I'ensemli ! uii
tableau, elle avait besoin de tournures breves et nettes qui '►ni
autant et meme plus que ne dise une prop, subordonnee. C'l't t It
lieu de se creer des genitifs et des locatifs absolus. Le preiiiii I - ttt
developpe du genitif adnominal (cf. les exemples cites dessiis) i ^
en grec et il est probable que ce fut Taboutissement — inconql i n
Sanscrit d une tendance indo-europeenne (cf. AV'ackernagf'l. A
ii. Synt., i, 292).
Mais avant le genitif absolu le vieil indien a deja eu le tenip^ ■ ' 'f'
creer le locatif absolu, dont la valeur concrete et adverbiale n
aide a fixer son application. Le genitif, cas granuuatical, etait ' n''
de nature (voir ci-dessus). II se pretait, avec son caractere coini’i s>'.
a la phrase poetique, toute en sous-entendus ; se inontrait u
utilisable, parfois dangereux pour la clarte. en prose. C'etait an-
d empechements a vivre. D’autre part la categoric de temps penet
de plus en plus le systeme de la langue. le genitif independant di'' i ^
quand et “ apres que (cf. tdsydlabdhasyn chez Delbriick, A.S.
s est trouve inutile ; des qu’on voulait souligner le moment de 1 ai f '
sans plus, le locatif abs. y suffisait parfaitement ; avec le parfiri "
passe, il iiidiquait 1 action secondaire achevee avant Taction princip.i'
avec le participe present — leur simultaneite. Au fur et a mesure i
developpement des procedes syntaxiques, cette seconde fonction eclc.
au genitif abs., mais le tour verbal cedant dans la langue de plus '■
plus au tour nominal, le locatif abs. avec son participe passe passif -
adjectif, convenait bien au systeme evolue, tandis que le genitif ab.'.
avec son participe present a valeur verbale s'est trouve du cote d* '
formes refoulees. Aussi son application devint-elle restreinte et le Ixu;
usage du temps de Panmi Ta reservee aux cas particuliers {anadara)
La langue plus negligee et plus vivante des epopees et des Purana>
(cf. les exemples chez de Saussure) n’y a pas apporte beaucoup dr
changement . le genitif absolu forme avec des participes moyeii''
ou des participes actifs des verbes indiquant Tetat {cint, pas et pareils)
nest plus qu'une tournure adverbiale, une expression toute faitc.
epourvue de vitalite et rare, car peu necessaire. Il n'a jamais etc
vivant a Tetat pur.
Dhola-Marura Duha : A Fifteenth-century Ballad
from Rajputana
Is the present conclusion original ?
By T. Grahame Bailey
rriHE story of Dhola and Marti is told in a stirring Rajputanl ballad
-L published in the Balabakhsh Rajput C'aran Pustakmala series.
It is reviewed on another page of this Bulletin.
The story of the poem is briefly this. Pingal, the king of Piigak
had a daughter called Maravanl ; Naj, the king of Narvar, liad a son
named Dhola. During a famine Pingal sought temporary refuge iti
Xarvar, where the two rajas betrothed their children to each other.
Some years after this Nal, reflecting that Pingal lived far away, and
that the journey to his country was perilous, married his son to
MalavanI, daughter of the Raja of Majva. In due time Pingal sent
messengers to call Dhola, but the wily MalavanI had them killed.
I Itimately Maravanl succeeded in getting a message delivered by
singers. Dhola was charmed by their description of his early hancee
and set out for her country. After some vicissitudes he reached her,
and they were married. On the return journey Maravanl died of snake-
bite and was restored to life by a jogi. She was nearly seized by a
Muhammadan chieftain, but was warned by a Gipsy woman, and
through the swiftness of her camel, which, like MalavanI's parrot,
liad the gift of speech, she and Dhola got to Narvar in safety. There
they all lived in mutual affection, an affection clouded once bv
a domestic disagreement. Each of the wives praised her own country
and decried that of the other. Dhola supported Maravanl, and this,
rather inconsequently, restored peace.
This episode, which forms the conclusion of the poem, strikes me as
unnatural and out of place. The story appears to end properly with
doha 653, which tells us of their settling down in peace, and .says that
it was God Who had joined them in this happy union.
Now when we think the poem has come to a suitable ending there
''tarts a sudden argument between the two wives about the merits of
their respective countries. Dhola supports Maravanl. whom he
obviously prefers to this other wife, and his one-sided attitude appear.s
to satisfy even MajavanI, whom he had failed to uphold. Once again
the poem comes to an end. The final words closely resemble those of
976
T. C. BAILEY —
cloha 653. The two conclusions are alike ; the sense and '■ il of
the actual expressions of the last three lines (dohas 673, 074' * the
same as in the previous ending, dohas 651 and 652. Not onh . but
three other dohas. Nos. 666-8, are almost letter for letter the le
dohas in an earher part of the poem. I have drawn attention hem
below.
After Dhola and MaravanI reached Narvar in safety, we it
(doha 651) Dholdu Narvar dviydu, mangal gdvdi ndr
uchav huvdu aydu ghare, harakhydu nagar apdr
Sdlhlcumar bilmdl soda kdmin sugun sugdt.
Dhola came to Narvar, the women sing songs of rei mg-
There was a feast ; he came home ; the city rejoiced ' ond
measure. Salh Kumar (i.e. Dhola) made merry with his '
virtuous and beautiful.
The next doha appears to end the story.
653 Mdravani ndi Mdlavani, Dholdu tin bhartdr
ekani nmndir rang ramdi, H jofl Kartdr.
iVIaravani and Malavani, and Dhola their husband lived jo idt
in one palace ; God had made their union. (This hen, uh
reminds us of Tennyson’s “ marriages are made in heave. )
654 tatkhan Mdlavani Icahdi,, “ sabhali kant surang
“ sagla des suhamnd, Mdrii des virang.
At that time Malavani says : “ Listen, charming hu>l ^ ■
every country is beautiful, (but) Marvar is insipid.
655 “ bdldu, bdbd, desrdu, pdni jihd kuvdh
ddhvrdt kuhakkard, jydu manasa muvah. ^
“ I would burn up, father, a land where the water is in wells. ■■ ''
at midnight there is a shouting as if people had died.
656 bdldu, bdbd, desrdu, pdni sandl tdti
pdni kerdi kdrandi pri chanddl adhrdti (v. 1. sicai).
I would burn up, father, a land with anxiety about water, uh'
for the sake of water, the husband leaves (the house) ut
midnight (v. 1. draws).
657 bald, Dhola, desrdu, jdi pdni kuven
kUku varand haththrd naht su ghddhd jen.
I would burn up. Dhola, a land where water is in wells, and uh?'*^
red-coloured hands do not draw it. (ghddhd, of doub u
meaning ; perhaps connected with H. kdrhm ; Pj- kad(lk>p' ■
Ks. kadun.)
DHOLA ilARURA DUMA 977
()r>8 baba, ma deis Mdruvd, sudhd evuMh
kandhi kuhdruu, siri ghamu, va'iiiu imujhi Thuldh.
Father, Thou shalt not give me (in marriage) to Marvar, to simple
shepherds, axe on shoulder, waterpot on head, to live among
(the people of) Marusthal (Marvar).
659 bdbd, imi deis Manivd, var kddri rahesi
hdthi kacoldu, siri ghamu, slcantl ya maresi.
Father, thou shalt not give me to Marvar, I will (rather) remain
virgin from a husband ; cup in hand, waterpot on head, I
shall die drawing water (or watering) ; (i.e. if I go to
Marvar).
660 Mdrau, thakdi desrai ek na bkdjdi ridd
ucdldii ka avarasanau, kdi phdkau, kdi tidd.
0 Maravanl, in your country, not even one difficulty flees away ;
there is either journeying (from the country), or lack of rain,
either hunger or locust.
661 jin bhui pannag plyana, kayar kantdrd rukh
dke phoge chdhrl, huchd bkdjdi bhiikh.
A country in which are (blood)-drinking snakes, and the trees
are thorns and thorny shrubs ; the shade is only dk and leaf-
less shrubs, and hunger flees by (eating) hfich (thorny plant,
the seeds of which are eaten).
662 pahiran-orhan kambald, sdthe purise nlr
dpan lok ubhdkhard, gddar chdti khlr.
F^or clothing and putting on (only) blankets ; water sixty puris
deep ; the people themselves wanderers ; milk (only) of sheep
and goats. (A puris is about four feet.)
Maravanl replies by running down Malva and praising Marvar.
663 vaiafi Maravanl kahdi " Mdrfi des siirang
idjd tdu sagld bhald, Mdlav des virang.
In turn (returning) Maravanl says ‘‘ Maru land is charming :
others indeed all are good, (only) Malva land is insipid.
664 bdlu, bdbd, desrdii, jaha pani sevdr
nd panikdri jhulardu, nd kfivdi laikdr.
1 would burn up, father, a country where the water has sevdr
growing in it ; {sevdr, Hindi skaivdl, a water plant) ; neither
companies of water-women, nor melody at the well.
665 bdlu, bdbd, desrdd, jaha phikiriyd log
ek na disdi goriyd, ghari ghari dtsdi sag.
978
T. G. BAILEY
I would bum up, father, a land, where the people un-
interesting ; women are not seen, even one ; in every ise is
seen sadness.
666 Mdru des ujpanniya, tihdkd kd dant suset
kujh bad gorangiya, khanjar jehd net.
This doha has already occurred as No. 4-57. where for >i, ">>fi
we have upanniya, a better reading. I assume it here.
Girls born in Marvar, their teeth are beautifully white : ti, > are
fair as young cranes, and their eyes are like those ot taib
(or are like wagtails).
667 Mdrii des upanniya, sarjydu paddhariydh
karvd kade na bolahl, mithd bolaniydh.
This is almost word for word the doha which we have already i I a'
No. 484. There the ending is -ydh, which is preferable.
Girls born in Marvar are straight as an arrow, they never ak
bitter words, they are speakers of sweet things.
668 des nivdnu, sajal jal, mithd hold loi
Mdru kdmini dikhani dhar Hari dlydu tdu hoi. "
This doha, with one word of difference, occurs as No. 485.
The land is low-lying (therefore fertile), fresh in water. k
people speaking sweet words ; MarvarF women (woineii
them) might be in the land of the south, but only il '
gave them.
Now Dhola speaks and favours MaravanF.
669 des surangdu, bhul nijal, na diyd dos Thaldh
ghari ghari cand-vadanniyd, nir carhdi kainldh.
The land is charming, (yet) the soil is waterle.ss ; do not attrihin
fault to Marusthal ; at every door are moon-faced girls, hk'
lotuses which rise to the water (or who a.scend to the watei
like lotuse.s).
670 suni, sundari, keld kahu Mdru des vakhdn
Mdravanl miliyd pachdi janydix janam pravdn.
Listen, fair one, how much shall I praise Marvar ? Since I niet
MaravanF I have regarded my life as fruitful.
Thi.s, while a charming compliment to his favourite wife, wa-
depre.ssing to the other. Reading the compliment we think
Browning's :
The purpose of my being is accomplished
And I am happy. I, too, Federigo.
979
DHOLA-MARURA DUHA
671 jhagrau bhdgau goriyd, Bholm purl sakhkh
Maru ruliyait kill, pdml priya parakhkh.
The quarrel of the fair ones fled away : Uliola supported
(MaravanI) ; MaraA’anI became liappy ; she had tested her
loved one.
67:1 Malav des vikhoriya, Mdru kiyd vakhdu
Maru sohagin thal sitndari sagun sujdy.
He decried MaHa. and praised Marvar ; Maravani. beautiful,
virtuous and wise, became fortunate.
673 jim- madhukar ndl ketakl, jim- koil sahkar
MaravanI man harakhiydu dm Dholdi bhartdr.
As the bee and the keord. as the koel and the plaintaiii. so
MaravanI's soul rejoiced in Dhola. her luisband.
We are not told how the other wife regarded the situation, and
the final couplet which follows is unnatural at this point.
671 dnand ati, uchdh ad, Narvar mdJie Dhol
xasnehl sayand tand kalimd rahiyd bol.
Much happiness, much feasting. Dhola in Narvar ; and the story
of those loAung lovers continued in this iron age.
The sense of this doha is the same as that of No. 653.
Ii3
ViJL vm. PART 4.
Phonetic Observations on the Brahui Language
By M. B. Emeneau
I N January of 1936 Dr. Ernest Mackay, director of the excavation
of the American School of Indie and Iranian Studies at Chanhu-
daro, courteously arranged that I should have the use of a Brahui
speaker and an interpreter for a number of sessions. The Brahui.
Dad JIuhammad by name, belongs to the Nicharl tribe of the Jhalawan
division of the Brahuls.^ Some confirmatory notes were made as well
with the aid of another of the Brahui speakers working at the
excavation, of the same tribe and village as Dad Muhammad. The
notes made were chiefly phonetic. The results in the main need not be
detailed here ; Sir Denys Bray’s account is entirely accurate for the
phonemic system of the language, except, it seems, in one small point
to be noticed later, and needs to be supplemented only by a few phonetic
observations, some of which I make in this paper. All examples are
written phonemicaUy unless indicated otherwise. Accents are not
written ; all polysyllabic words quoted have a stress accent on the
first syllable, except [ka?e:k], accented on the second syllable and
treated in the paragraphs of Bray, i, referred to below when the word
is quoted.
The phonetic system of Brahui is, on the whole, simple and straight-
forward, and the phonemes show few variants. The fricatives [x]
and [y], for example, Bray's lA and are always made in the velar
position, rather far back towards the uvula, but not involving any
trilling of or friction against the uvula ; they are not advanced in
position before or after front vowels. Examples ; [xon] " eye .
[xa:xor] “ fire ”, [troxios] “ crack (especially chap on hand or foot
caused by cold), [xiisun] “red", [xe;r] "behind , [li:x] "neck’,
[ti:xta] “ put it down ”, [se:xa:] “ shade ”, [patfx] " bark of tree ” ;
[beiy] “ knead (imperative) ", [liyi^] naked ”, [your ke:s] ‘ think it
over ! ” [i: piryadta] “ I broke it ”, [zaym] sword ”.
The two tremulants, Bray's r and r, are respectively [r] and [r].
The former is a voiced trill of two or three or more taps made just in
front of the alveolar ridge. Examples : [brfl;^u:i:] Brahui , [^ikj
^ See Bray, ii and iii, p. 4. References will be made to : Bray, i ~ The Brahui
Language j part i. Introduction and Grammar, by Denys de S. Bray, Calcutta,
Superintendent Government Printing, India, ItW. Bray, ii and iii - The Brahui
Language, part ii. The Brahui Problem; part iii, Etymological Vocabulary, bv
Denys Bray, Delhi, Manager of Publications, 1934.
982
M. B. EMESEAU —
“ jump (noun) ", [irat] “ two ", [ma:r] " son . [i: piryaittaj i ikr
it ”, [o: trn^ra;] " he choked (in drinking water so that the w.i’ mit
out of his mouth) This [r] is also lengthened and made wiri • "r
six or more taps (written by Bray rr). This [r:] is phonemicall}' ' ; ’ ' ''t
from [r] ; e.g. [ora;] “ which ? " [^ar:a:] " (he) tore (it) ‘ i
is final the last few taps tend to be unvoiced, e.g. [bir:] "
narrow transcription might be written [birf]. The other tn ' n''
[f] is, in Dad Muhammad's speech at lea.st, a single-flap .'ioun l i'h'
in the same position as [r], with the fully-voiced downflap
prominently when the phoneme is final or before another coti'^ a
Examples: [va;r] “hedge’’ contrasted with [va:r] ■’turn,
[ro:f] “calf”, plural [ro:fk], “dismount (imperative)
trasted with [^ar] “takeaway (imperative) ”, [i: tare;tta] I en
[?arde;] “ every day ’’.
The two liquids, Bray’s 1 and IJh. are respectively [1] and [H-
made just in front of the alveolar ridge. The voiceless lateral i.'!' a' ”■
panied by a coasiderable amount of friction. Examples: [^al] ’
[^<4] “ seize (imperative) ”, [?4] “ fever ”, “ a fly ,[’id] ^
for a person’s authority ”, [xal] “ stone ”, plural [xalk], [xal] P'*'
[i; xalkutta] “I beat him’’, [xojta] “beat him", [pa4] "an*
[me:}] “ sheep ”, plural [me:}k], [3a:laj “ hail ”, [te:}a:mba:] la. >
scorpion
In forming the dentals [t], [d], and the groups [nt] and [n^]
tip of the tongue touches the back of the upper teeth with a cert,
amount of contact of the edge of the tongue all around the backs of '
teeth, [n], when not followed by [t] or [d], has its contact somewl
further back, probably at the roots of the teeth, [t], [d], and the group'
[nt] and [nd] are formed by contact of the tongue-tip with the ah col. n
ridge, not, as in the English sounds, by contact of the blade with tin
alveolar ridge, [n] is said by Bray (i, p. 28) to occur in some words not
in the neighbourhood of [t] or [dj, but I have no examples of it and am
unable to say anything about its point of articulation. Only a fe"
examples need by given to supplement the material contained in th<
preceding and following lists of examples: [tu:] “month ,
“ fat ”, [ant] “ what ? ” [ko;nd] “ cave ”, [ko:nt] “ wool rug.
drugget ”, [ko:ndo:] “ fowl-ticks
The phoneme given by Bray as h is probably to be divided into
two : a voiceless aspirate [h] and a glottal stop [?]. The former wa^
found in a few words only, always intervocalic : [a:haf] “ mid-
summer ”, [baha:] “ price ”, [baha:na] “ excuse ”, [^aha:r] " wild
PHONETIC OBSERVATIONS ON THE BRAHUI LANGEAGE 983
animal Although no words could be found differentiated from
these merely by the occurrence of [^] instead of [h], yet such words as
[ta’Q-.r] “ dark [ma?a;lau] “ soon, in good time ", make it impossible
to set up [h] as a positional variant of [?].
[^] occurs initially in all words in which Bray writes h. including
those for which he gives variants without h. except, it seems, the word
[ara;] “which?" Examples: [^aine;?un] “just in that way",
“ hajji ", [o; ^eista] ‘‘he brought it " [^impo] "'do not go “.
[’is] "ashes”, [’u^] ‘'camel", [’ust] “heart". [’o:r] ‘‘finger”,
plural [’o;k]. Other examples are to be found in the preceding and
following lists.
Examples of intervoeahc [’] are : [ka’eik] “ (fire) is dead " (Bray.
1. §§ 203, 206), [ke:’aik tam;a:] "(he) stumbled and fell”, [ni: ante
xwQ;’isQ] “what do you want ?" [oi’un] “in that way", [pei’a;]
" (he) entered ", [pii’un] '■ white ", [sa’i: of-.at] “ I don't know ".
In absolute final position and before a consonant [’] is very
weak and hard to hear, being somewhat masked by a very short but
fully voiced re-articulation of the preceding vowel. Examples ;
(da;’] “complaint", [^o:’] “treachery", [de;’] ‘‘region", [mu’]
"ahead", [jo:’] ‘‘9"; [mai’ri:] “riding camel", [mei’r] “kind-
ness", [mi’iijQt] ‘‘wages", [mo;’Je‘.f] “descent". A comparatively
large number of words has the vowel [a] preceding [’]. The vowel is
fronted and raised to a position between [se] and [e] ; the words are
given here with the phonemic transcription [a]. Examples : [du’j
10 ’> “abode", [gu’] “jibbing"’, [no’] “no", [ra’] “side,
river-bank" (cf. [ra’at] “ by the bank’’, with [a] unmodified), [tra’j
" sudden start ", [za’j “ flock of kids and lambs " ; [ka’v] “spleen .
[ma’r] “ dowry settled on wife by husband ’, [pa’ro] “ sentry-go ,
[pa’re:z] “avoidance", [sa’t] “jewellery", [Ja’d] “honey", [Ju’r]
“ village ", [ta’ti: ’urak] “ look inside ", [zo’r] “ poison .
The diphthongs, Bray's ai. ei. ai, au, are respectivelv [ai]. [ae].
|Q;i], [au]. Examples : [aiuo:] ‘‘ to-day", [Jair] ” song [ba xairat]
" welcome ! " ; [bae] “ grass ". [antae] “ why ? [sae] “ three " in
[bi:st o; sae] “ 23 ". [si: o: sae] 33 ", etc., [saejombe:] ” Tuesday ",
fjaeta;n] ■ devil ; [da;iska;] ‘' up to now ", [i; ba:ife;t] “ I lost at
gambling " ; [lauz] “ language [kulau] “message'’, [’au] “yes",
i’aud] “ small tank ". The word for “ black"’ which Bray gives as
niauH was pronounced bv my informants [mo:n], homonymous with
[ino;n] “face". So also [do;^] “in this way " was given for
Bray s daun.
Ein tiirkisches Werk von Haydar-Mirza Dughlat
Von Ahmet-Zeki Validi
FTNTER den ostturkischen Mss. in der Martin-Hartmann-Sammlung
der Berliner Staatsbibliothek gibt es ein in Versen (mutaqarib)
verfasstes Merk namens Gihan-Name (Ms. Oz. Oct. 1704). das in dem
Verzeicbnis Hartmanns {MSOS, vii, 1904, Bd. ii. S. 10, N. 83) nur als
.. Erzablung iiber Firuz-Sah bezeichnet ist. Als icli 1924 diese
Sammlung nacb dem Vorschlag von Prof. Weil katalogisiert babe,
stellte ich fest, dass es sicb bier um ein unbekanntes Werk von Haydar-
Mirza Dugblat bandelt. Die Hs. bestebt aus 125 Seiten im 16” und
die Abschrift ist von eineni gewissen MoUa ‘Umar ibn Molla Hagi
am 16. 6umadi ii, 1229b. (5. Juni, 1814) vollendet worden.
Der Inbalt des Biicbleins ist eine Erzablung. Sie lautet :
Prinz Firuzsah, der einzige Sohn des Konigs Sabsewar und der
Konigin 6iban-Banu von der Stadt §ehri-STmIn(Silberstadt) verliebt sicb
im Traum in ein Madchen aus unbekanntem Lande. Mit dem Gebeimnis
des Prinzen ist nur Firuze-Ray, der Sobn des Wezirs Selim, vertraut.
Der infolge der Sehnsucht nacb dem unbekannten Madchen sehr
bekummerte Prinz wird auf einer Jagd durcb eine Gazelle in die Wiiste
gelockt, er yerirrt sicb und, die Gazelle verfolgend. kommt er zum Schloss
Gubstan-i-iram, dem Wohnort des unbekannten Madcheas. Sie ist die
Prinzessin Perizad, die Tochter des Konigs Tag-Baht und der Konigin
Mabru aus der Stadt Sebri-Zerrin (Goldstadt), dem Lande der Feen.
Die Bedingung fiir die Heirat, die schlafende Prinzessin durcb ^vunder-
volle Marcben dreimal aufzuweckeu, wird voin Prinzen glucklicb erfiillt.
Ihm kommen zur Hilfe der Throii, die Weinkaraffe und die Kerze, die
>ich bei der scblafenden Prinzessin befinden.
Zuerst erzahlt ihm der Thron. wie er. ein Holzstiick aus dem Niltal.
nacb langen Erlebnis.sen zum Thron geworden ist. Demgegeniiber
erzahlt der Prinz die Erlebuisse eines Holzcs, wie es durcb Wundertaten
ernes Gelebrten, eines Goldschmieds und eines Schneiders die Gestalt
vines scbonen, lebenden Madchen annahm. Jetzt entsteht die Frage ;
wem von diesen dreien gehort das Madchen ?
Zum zweiten erzahlt die Weinkaraffe iiber ihre Erlebuisse, wie sie
als ein Steinfels auf dem Alburuz-Gebirge am Heerweg der grossen
Eroberer Darius, Alexander u.a. lag imd wie sie von einem Steinhauer
gesprengt, weggeschleppt, behauen, gescblifien, endlich zu einer
11 einkaraffe geworden, und bis zu den Lippen der Prinzessin gelangt
1 st. Demgegeniiber erzahlt der Prinz von einem scbonen Madchen, das
von einem Damon entfiihrt war, von drei Personen : einem Fiirsten,
986
A.-Z. VALID!
einem kriegerischen Heklen und einem Steinschleuderer gerettet \' 1' .
Die Frage ist nun : wem von diesen dreien gehort das Madchc
Drittens erzahlt die Kerze iiber ihre Erlebnisse, ivie sie at- la;'
als ein Schaf zuerst gcschlachtet und dann zur Kerze der Priii 'in
geniacht wurde. Demgegeniiber erzahlt der Prinz von zwei Yerln n.
die nachdem sie von einem Rauber enthauptet waren. von i m
Reisenden wieder ins Leben zuriickgerufen wurden. Dabei er
wurden bei der Herstellung ihrer Gestalt ihre Kopfe irrtiimlicher i'e
verwechselt. Die Frage lautet : ivem von diesen zwei verln n
gehort das Madchen ?
Um auf alle diese drei Fragen zu antworten, unterbricht 'if
Prinzessin ihren Schlaf. Der Prinz bekommt die Hand der Prinze' i :
er lasst auch seinen Freund, den Sohn des Wezirs, Firuze-Ray ii u '
Sehri-Zerrin kommen, wo auch er ein Madchen namens Xazenin er' ^
Nach den Hochzeitsfeierlichkeiten in Sehri-Zerrin fuhren Prinz * i
Prinzessin Perizad nach Hause, wo Firuzsah von seineni alten ^ . ' r
und den Untertanen ,, von Tiirk und Tagik nach dem .. Tiiz;
empfangen wird. Nach dem Tode des Vaters kommt er auf den Th; "■
Nach hundertfiinfzigjahriger gliicklicher Herrschaft stirbt zum gro> a
Kummer des Firuzsah die Kbnigin. Ihr Leichnam wird von k- '
nach ihrer Heimat Sehri-ZerrTn gebracht und an einem unbekaiint' '
Orte heimlich beerdigt. Firuzsah durch den Verlust seiner gelieb' i
Frau tief betroffen und voni Gram niedergeschlagen, kann sich ni '
mehr mit den Staatsangelegenheiten befassen. Er schickt seinen Soi' ■
den Kronprinzen Farruh-Sewar in die Sehri-Zerrln, damit er sich in -
die Grabstatte seiner Mutter erkundige ; aber alles ist vergebens. D; ”
Wezir Firiize-Ray schlagt dem Konig, der die Staatsfiihrung vernai
lassigt, vor, auf den Thron zugunsten seines Sohnes Farruh-Sewar /•
verzichten. Er niinint den Vorschlag an. Er begibt sich nachher vi
Verzweiflung auf die Jagd in die M’iiste, wo er wieder dieselbe Gazcl •
trifft, die ihn friiher zum Schloss Gulistan-i-lram gefiihrt hatte. i'''
fiihrt ihn jetzt zur geheinien Grabstatte seiner Geliebten, wo auch > a
seinen Tod findet.
Tm ganzen Buch ist der Verfasser nirgends namenthch erwahnt. lu
seiner Vorrede (S. 2-11) und im Schlusswort (S. 121-125). die ich hici
beilege gibt er aber das Datum der Abfassung und die Schilderung dci
Umstiinde, unter denen das Buch geschrieben wurde, welche verratcn.
dass der Verfasser kein anderer war als Haydar-Mirza. Das Buch n't
vom Autor selbst Gihdn-Natne betitelt und im Ragab des Jahres
939 (in Buchstaben ; oXJ>- ^s), d.h. 27, i-25, ii, 1533, im Pferdejahr
vollendet worden (S. 124, v. 10-12). Diese Tage, in welchen er offenbar
nur den Schhissteil geschrieben hat, gehorten, wie aus S. 124, v. 1-
S. 125, v. 3, hervorgeht, zu der schwersten Zeit seines Lebens. Er
konnte sich damals nur durch die Erinnerungen an das Vergangeue
EIN TUEKISCHES WEEK VOX HAYDAR-MIEZA DUGHLAT 987
Trosten. Nach den Ausfiilirungen in der Vorrede und ini Nachwort
;st der Verfasser beim Chaqan oder Chan Osttiirkestans (Sa‘Id-Chan,
den er aber nirgends mit Namen nennt) so beliebt, wie Ayas bei Mahmud
von Gazna (S. 8, v. 6) ; auch sein dichterisches Pseudonym war Aydz
(S. 122, V. 4). Er war mit dem Chan schon im ,, Vilayet “ (S. 5, v. 3),
■ lass in der Sprache der damaligen Tiirken und Tagiken Turkestans
I .. bebaute Bezirke und Stadte von Mawerannahrim " heisst,
' zusammen. Dann haben der Chan und er (der Verfasser) einen Feldzug
jxegen Andigan unternommen (S. 6, v. 10). Von hier aus sind sie mit
ilirem Heer weiter gegen den Herrscher von Kaschgar, Ababekir-
IMirza gezogen, um den Thron, der den Vorfahren des Chans immer
gehbrte, zuriickzuerobern. Kaschgar und Chotan wurden besetzt.
Ababekir wurde in die Flucht geschlagen und auf dem Wege nach
Tibet festgenommen und getotet (S. 6, v. ll-S. 8. v. 2). Dies sind also
die Ereignisse, die Haydar-Mirza in seinem Tarikh-i-Kashidi (transl.
by Denison Ross, S. 241 f.. 247 f., 310-325) so ausfiihrbch schildert.
Dann erziihlt der Verfasser Uber eine Abordnung aus Badahsan. die
gekommen sei, um sich liber das traurige Schicksal ihres Landes zu
iieklagen. dass namlich Badapan sich in einem Zustand der Anarchie
befinde und. falls es nicht von SoitendesChan von Kaschgar besetzt wiirde,
in die Hiinde der Ozbeken iibergehe. Der Chan entschliesst sich, dieser
Bitte nachzukommen. Er ziehe mit seinem Heere nach Badahsan,
wobei er den Verfasser mit einer Vorhut vorausschickt. Der Verfasser
kommt nach Badahsan und dort erinnert er sich, dass er hier bereits
19 Jahre vorher gewesen war. Er fiihrt weiter aus, dass von seinen
damaligen Freunden in Badahsan niemand mehr da war, und dass er in
den schlaflosen Nachten diese Erzahlung in Versen niedergeschrieben
babe (S. 9. v. 10-S. 11, v. 5). Dies ist die zweite Expedition des Sa'id-
Chans im Jahre 1528 von Yarkend aus nach Badahsan, wo Haydar-
Mirza tatsiichlich mit einer Vorhut (IManghalay) vorausgeschickt worden
war (vgl. Tarikh-i-Eashidi. S. 387 f.). Sa'Id-Chan und Haydar-Mirza
Haben bei dieser Unternehmung ,, QaTai-Zafar die Hauptstadt von
Badahsan drei Monate lang ohne Erfolg belagert. Von diesem Feldzug,
richtiger Streifzug, Sa’id-Chans erwiihnt auch Babur als von einem
undankbaren Vorgehen seines Neffen, demgegeniiber er so viel Gutes
in den schwierigen Zeiten seines Lebens getan habe (Edition Ilmenski,
501-3). Auch die Bemerkung, dass er (der Verfasser) 19 Jahre vorher
in Badahsan unter seinen Freunden gewesen war, die nicht mehr da
sind, stimmt mit den Einzelheiten des Lebens Haydar-Mirza' s iiberein.
Tat.sachlich war er vom Ende des Jahres 1508 bis zum Ende des Jahres
988
A.-Z. VALIDI —
1509 in Bada^n, er weilte damals in Qalai-Zafar bei seinein ( i 1
dem Temuriden Han-Mirza (s. Tarikh-i-Rashidi, 215, 221). di ’ I'l
Jahre 1528 nicht mekr am Leben war.
Der Verfasser hat das Schlusskapitel spater, wie schon erw.' r.
im Januar-Februar des Jahres 1533 vollendet. Zu dieser Zeit war ! '
dar-Mirza in Kaschmir. Sa‘id-Chan hat ihn bei seiner Expedition ii
Tibet von Tubra aus mit einem Heer von 2000 Mann nach Kas( ' ,;r
beordert. Der Chan selbst zog nach Yarkend ab und starb unter'. '
am 9. Juli 1533. Haydar-Mirza ist nach Kaschmir im Winter und z' u
im Gumadi II (beginnt 26, xii, 1532) gekommen und bis zum Sau il
(Mai 1533) dort geblieben. Den Schlussteil des 6ihan-Name vollcm .
er also unmittelbar nach der Ankunft in Kaschmir.
Das Werk ist geschichtUch und sprachlich sehr interessant. F* i
Verfasser sagt am Schluss S. 124, v. 3, dass die Erzahlung seinen eiget. i
Erlebnissen entsprach. Damit meint er offenbar das Hauptmom
in der Erzahlung, wonach ein Herrscher, der nicht mehr den 8t.
fiihren konnte, gezwungen war auf seinen Thron zugunsten sen
Sohnes zu verzichten und sich in die Wiiste zu begeben. Daraus ka*
man ersehen, dass auch sein Herrscher Sa‘id-Chan die Expedition na ''
Tibet (1532-3) nicht ganz freiwillig unternommen hatte. Sa’id-Clr
hat wahrend derletzten Jahre seiner Regierung die Staatsfiihrai .
vernachlassigt (er trank viel) und dabei hat sein energischer v8o*'
Abdurrasid offenbar immer mehr Ansehen gewonnen. GeschicF*
lich gesehen ist auch die Bemerkung interessant, dass der Feldzu
Sa‘Id-Chans im Jahre 1528 ein Zuvorkommen vor der Eroberuu
Bada^ans durch den Ozbeken bedeutete.
Das Werk zeigt ausserdem, dass Haydar-iVIirza im Tiirkischen eii
ausgezeichneter Dichter gewesen ist. Seine Sprache ist, wie die seiiiti
Zeitgenossen Babur und Saybaq-Chan, einfach ; die Schilderungen
sind klar und lebendig. Er hat eine Erzahlung, urspriinglich vielleicht
eine indische,^ durch die von ihm eingeschobenen Bilder aus dem Lebeu
der Herrscher Tiirkestans zu einem tiirkischen gemacht. Untertanoii
des Konigs Firuzsah sind „ Tiirken und Tagiken die Zerenionien
sind in „ Tiiziik “ (Traditionsgesetz) vorgeschrieben. Besouder'
interessant und lebendig ist die Schilderung der Jagd (S. 26-27)
Dort dirigieren die uns in der Ge-schichte der Djagataier und
Temurs wohl bekannte ,, tawaci’s die Sohne der Bek s
bilden bei der Jagd einen besonderen „ quram “ (Gruppe, Abteilung)
‘ Vgl. z.B. P. Kretschmer, in WZKM. B. 37 (1930) S. 15 f.
EIN TURKISCHES WEEK VOX HAYDAR-MIRZA DUGHLAT 989
Interessant sind die Termini und Ausdriicke tufqal^'. tiiziik
girgeledi'", „ girgadia qalmaq'", „ yibardi iitii her sari bes qonaq‘'.
Merkwiirdigerweise sehen wir in der djagataisclien Sprache Haydar-
Mirza's, so wie in der des Saybaq-Chan Spuren des Westtiirkischen, z.B.
nluh, durur, statt bolub und turur, dazu die Auslassung der Yokale.
8aybaq-Cban hat bei seinem Feldzuge gegen die Chane der Qazaq
(Kirgizen) den Alexanderronian des westtiirkischen Dichters Ahmedi
bei sichd Es ist moglich, dass Sa‘id-Chan und seine Umgebung sich
auch fiir die westtiirkische Literatur interessiert haben ; genau so
wie wir dagegen den Vers des westtiirkischen Dichters Fudidl :
j'j
Z.B. mit dem des Sa'Id-Chan :
j (^^***^ sdij 'AI****. 'AL-j ^
vergleichen konnen. Babur in ludien, Saybaq-Chan in ^\ esttiirkestan.
8a'id-Chan und Haydar-ilirza in Osttiirkestan waren, trotz der
politischen Gegensatze, Angehorige ein und derselben Schule in der
djagataischen Literatur und standen oifenbar auch mit den
literaischen Strdmungen der Tiirken Westasiens in Fiihhmg.
^ Dariiber in ,, Mihmanname-i-Buhdra ” von Fadlullah ibn Ruzbehan al-l?!fahant,
Hs. der Xuri-Osmaniyeh in Istanbul, N. 3431, f. 131b,
' Diwan-i-FudulI. Istanbul, 1328, s. 143.
^ Bei Denison Ross. Tarikh-i-Rashidi. S. 138 ist die-er Vers, wie auch die andere,
diisgelassen.
The Arabic Theatre in Egypt
By Xev[ll Barboi r
PART III
The Plays
'T’BASSLATIOXS and Adaptations. — As has been stated in the
preceding article, most of the plays produced in Egypt from
1875 to 1914 were translations or adaptations of European originals.
Shakespeare was amongst the linst to' be utilized, and a version of
Othello, with the title -ClaJI j\ was played by al
Qurdahi. It is difficult to form a clear idea of what these early
performances were like ; but it is obvious that the very amateurish
production, the frequent oriental songs and the changes to suit
local taste must have resulted in something very different from the
productions of Shakespeare that are current in England to-day.
Shailffi Salama used to take the part of Romeo in a translation
of Romeo and Juliet _• j' made by
Xajib al Haddad. Hamlet was played by both al Qurdahl
and ^ailffi Salama in a version made by Tanyfi.s 'Abdu. After
Jurj Abyad s return from Europe, he appeared in several plays
of Shakespeare in translations bv Khalil Mutran.'^ These are highly
praised, notably Macbeth Handel and Othello
(JJac).2 Another version of Hamlet, with the title was
published by Sami al Jardlnl in 1922. A version of the Tempest
by Dr. Abu Shadl, was published in 1929, and the same
play has been produced by Fatima Rushdi's company in a version
^ Specimens in the MvS. of Taiifiq Habib, pp. li6-lS.
Khalil Mutriln rejects 4 jJ| as the original of the name •* Othello on the
grounds that this name is never used by Moroccans. He suggests i Uc. as an
'Affectionate diminutive from ( ' I'nadorned "), suitable to a <lark-skinned
Moroccan whose mother, at anv rate, wa'^ probablv a negro slave. " al Akhbar,*’ 26th
April, 1916.
992
X. BAEBOUR —
by Ahmad Eaml. Other Shakespearean plays produced in ent
times are the Taming of the Shrew and Julius Ccesar in a 'ion
by Mahmud Hamdr. A translation exists of King Lear, thou the
play has not been produced. Mohere was also amongst the i r to
attract attention, though he does not seem to have been as p ilar
as Shakespeare. A1 Qurdahl played L’Avare (al Bakhil) in a tran ion
made by Najib al Haddad, and a v^ersion of Le MMecui ntahi '//,
with the title at Tabib was made by Iskandar Habqall {]). ue
exists also a volume, printed about 1900, containing four pla of
Moliere in colloquial verse, by ‘Uthman Bey Jalal. Other aut > ir<
who were put under contribution include Corneille, Racine. ^ ‘ or
Hugo, Dumas, and in more recent times Rostand, Bourget, and i uv
other modern French playwrights.*
The second stage was the adaptation of modern pieces u ui
Egyptian setting. This is done in its simplest form by moving the ; ue
from London to Cairo, by the metamorphosis of John into Muhanu uh
Mary into Fatima and the making of any other slight alterations w .c h
seem imperative. As it is at present impossible for a foreign dran
to enforce any claim for royalties in Egypt, such plays can be proc d
cheaply ; were they more expensive the Egyptian theatre coul' 1 ^ 't
afford to present them. Sometimes the name of the original aui u
is mentioned, sometimes it is not. A piece entitled BayyumI Etc u.
which is simply a translation of the well-known French play Le i ■
Lebonnard, was produced by the Ramses Company in 1932—3 : a
this case not only was no mention made of the original author, but
piece w'as described in the programme as “ from the pen of Hasan i
BarudI, a play Egj’ptian in action, Egyptian in language, Egyptiu
in its circumstances, Egyptian in everything The adaptations m
in general well done ; a tolerably Egjqjtian atmosphere is often succes-
fully substituted for that of the original. Special mention should tc
made of the adaptation of Mr. Knoblauch's Kismet, played by Fatiin i
Ru^dl's company under the title “A Night from the Thousand
Nights (“Laila min al Alfi Laila ”), one of the most enjoyabi ■
^ The following translations have appeared in a series published under the auspi* ^
of the Ministry' of Education (1932—3). Shakespeare “ Al Malik Ur,” translated b’
Ibrahim Ramzl ; “ Tarwid an Namira ” {Taming of the Shreiv), translated by IbrahiR'
Ramzi ; Moliere, Tartuf ** ; Ghosts An Enemy oj the People ; Cornell!'*
Cinryiy translated by Khalil Mutran ; Victor Hugo, Hernani ; Banville, Gringoire.
“ Programme of Ramses Theatre, undated. I have been told that the phrase ‘ froi'*
the pen of’’ should be regarded as an indication that the piece is not original ; but
it seems unlikely that the ordinary playgoer would so understand it.
THE ARABIC THEATRE IN EGYPT. Ill
993
pieces given in Cairo. ^ The work of the adapter, Mahmud Bairain at
Tunis!, partly consisted in pruning the excess of “ Eastern ” verbiage
and imagery with which the English author sought local colour. The
success of the performance was then due to the admirable acting of
Aziz ‘Id and Fatima Ru^dl in the parts of the Beggar and his daughter
Xajaf, combined with the excellent use the adapter has made of the
Egj’ptian vernacular. As an example of the latter, we may take the
opening of the third scene of the first act, where the Beggar's daughter
seeks an excuse to get rid of her duenna for a few minutes. The
English version reads : — •
Marsinah : The sun grows hot.
Xarjis : How's thy border ? Will it be done by noon prayer ? I
promised it to the merchant.
M. : I hear, Narjis, I hear. Hast thou any yellow wool ?
N. : Yellow ? Yellow ? Did I not give it to thee erstwhile ?
M. (hiding the yellow wool) : 'Twas red thou gavest me.
N. : By the life of thy youth, 0 Marsinah, 'twas yellow.
^I. : Look thyself. Thou seest I lack it to finish the pattern.
N. : Alas ! MTiat’s to be done ? What’s to be done ?
M. ; Run to the wool market, 0 good Narjis.
N. ; All the way to the wool market ?
M. : ’Tis none so far for one as sprightly as thou, 0 sweet Narjis. Thou
didst promise it to the merchant — ^remember !
N. : I could have laid an oath with the All-seeing there was yet another
strand of yellow.
M. : Couldst thou in sooth ?
N. : AVell-a-day ! There's nought for me but to go. We must finish
the work or the money's lost . . . And 0 Marsinah ! No looking
out of windows or over walls.
M. : By Lady Fatima’s life of light ! What dost thou suppose ?
N. : Think of thy father. Thou knowest how he fears for thy safety.
Was not his first wife stolen ? His son slaughtered ? Art thou
not the last of his race ? Is not thine own mother in the tomb of
eternity 1 I tell thee, should one folly on thy part reach thy
father’s ears, ’twere the undoing of us both.
N. : Fear nought, 0 dear Narjis.*
^ An abbreviated and unsatisfactory version of Kismet in the classical language
has also been played, e.g. by a touring company under ‘Abdullah ‘Ukasha.
Kismetf E. Knoblauch, Methuen, 1912, pp. 42-3.
994
X. BARBOCR —
In the adaptation this becomes : —
i;'
^ d tKU* J' J-^' ^
tjjX . . . . aJ li ? ^ U dj-i
I 1*1 L*
Vl ) 3jj( jJiLljl (^d[)
’ • • J--^ jj Ojddd) _^S\
4)* I j^>- <>-^iall
• • • (^7-^;®“ ^ ^ (<l2^l~)
•V: (j^\. <^4 ^ ^ -> <^* J->Vl ^
Aj Luo 1 ) 1 3
.... Ajl_*jo
VU- tUia-j (3^' .... <>^3* ^ . . . tU) d^
AJ 3 3
J— -J a «L1^! ^ ^«ljd I
CT*^' -)*■ d ^ »y. '*.J ^ ^ «Ir*
Aib jj
* = • 1 '
o I J-
* ^ for future tense.
= ' k .f\ u.
THE ARABIC THEATRE IX EUYPT. Ill
995
**— ic- ^****"^ C—
L_j 1 3 3 ‘iU-A
aJIaaJ _)
'iiAc. jL^ o j^j -5 ^ (^-j) j-J
jT' ^y* * > _i^) — -5 ^~-^‘*'^ ij"^’'
(•^aI-' J ^a li J ^1 1 LwJl
It canuot be doubted that the rather stilted English of the original
has gained inmiensely in the translation into the lilting and expressive
Egi'ptian colloquial.
Revue . — As has already been mentioned, the Egyptian populace
has a special talent and liking for a type of entertainment called
■■ Revue ” or “ Franco-Arab " or " Operette “. In this type of enter-
faimuent the story is little more than an excuse for introducing popular
comedians in their favourite situations. Imitations of foreigners talking
Arabic are always popular, in particular the Turkish aristocrat, the
tlreek grocer or cabaret keeper, the Maltese pedlar, and the English
tourist talking with his dragoman. The Azhar Shaikh also conies in for
his share of mockery ; and I have seen represented eight Bishops of
an Eastern rite in full robes who, after being summoned to interpret
a royal dream, left the stage in a burlesque danse de ventre ; this was,
however, greeted with hisses by a part of the audience. The pieces
played by ‘Ali al Kassar are of this type : one of the more original
is entitled Egjqit from the Hijra until to-day The opening scene
is laid beside the Great P\Tamid at sunrise. "All al Kassar is discovered
m the guise of an Arab soldier who has been sleeping since the time of
■-fmr ibn al ’Asl. He is awakened by a lady with a papyrus staff, w'ho
personifies Egypt. She relates to him succinctly, in Classical Arabic,
the history of Egvpt from the time he fell asleep until the present day.
- Erom >[.S. kindly lent by Mme Fitinui Rushdi.
VOL. viir. part 4. ^
996
N. BABBODK —
As she reaches each period, the scene changes and a typical epi'i >t
that epoch is represented, passing rapidly from a half-serious opt u
to burlesque. The nine scenes represented the times of the r.i
Mu‘awiya, the Abbasides, the Tulunids, the Ikhshidite dynasts
Fatimiyyun, Salah ad Din, the Mamelukes, Napoleon, and finall; e
reign of King Fu'ad, typified by a beach scene at the Casino ti
St efano at Alexandria. This performance lasted one and a half ii' -
being followed by varieties.
An example of Egyptian operette in its highest form is avail. '
to the reader in the entertaining and poetical piece called “ The To' d
Diamonds” (“al ‘Ashra at Tayvdba ”), which the late Muhanin 1
Taimur constructed on the theme of “ Bluebeard
Melodrama. — The responsibility for the introduction of this f i
of entertainment to the Egyptian stage seems to fall on Yusuf Y a!
who is its chief exponent and himself author of some of these pn ■
A S3mopsis of the most popular of them, Auldd al Fuqara will sui.
to give an idea of the sentiments and construction of the rest.
Act I. — A rich Pasha, living in Cairo, maintains in his house ; -
impecunious brother and the latter’s family. The Pasha’s son sedui
his girl cousin and then abandons her to make a match indicated . '
him by his father. To get rid of the girl cousin, who is expecting .<
child, the Pasha marries her off to another poor relation, an hoin
fellah from the provinces.
Act II (Some months later). — A girl child having been born, ti,
honest fellah discovers that it is not his. He is counselled by his fatl'
to submit humbly (“ We are poor people ” — auldd al fuqard ). 1 1''
Pasha’s son arrives at this moment and proposes to carry on hi-
relationship with his cousin. Her brother comes in and learns the truth-
The Pasha arrives and there are general recriminations. Finally tin
brother snatches a gun belonging to the Pasha’s son and wounds him.
Act III (Fifteen years later). — The brother has come out of prison
and has taken to cocaine. The honest fellah has spent his time searching
for his wife, who has concealed herself with the child on account of tlm
scandal. The wife and her daughter, called Bamba, are working m
a low bar, in which this scene is laid. Bamba is very imhappy ;
Greek proprietor of the bar threatens to dismiss her for refusing to
make herself agreeable to a drunken client. Her uncle, the cocaine
* “ -Al Masrah al (vol. iii, Mu’allafat Muhammad Taimur). Cairo, 1341,
pp. 255-352.
* Colloquial.
THE AEABIC THEATRE IX EGYPT. Ill
997
fiend, arrives to buy cocaine from the Greek proprietor. The lionest
fellah also arrives, still searching for wife and child, and bv chance sits
at a table with Bamba, not knowing that she is his own daughter for
whom he is searching. There also arrives a young effendi called Ra'uf,
son of the seducer of the first act, grandson of the Pasha ; he is still
at school and has fallen in love with Bamha. having no idea that she is
his half-sister. She confides to him that she is unhappy because she
was married to a brute who was the cause of her taking to this life,
and because she is suffering from syphilis. A little later on in the
evening Bamba, in order to get money to pay for treatment for her
illness, steals her father's pocket-book, without, of course, realizing
that he stands in any other relationship to her than that of a casual
client. The theft is discovered and the police arrive ; by a remarkable
coincidence the officer in charge of the police is the father of Ra'uf
and Bamba, the seducer of the first act. The relations now recognize
one another in a series of heart-rending scenes. Incidentally, this act
introduces Yusuf Wahbl's well known presentation of a drug fiend,
a scene where a comic English tourist and his wife are brow-beaten
by their dragoman, and another in which an Italian girl, having
been sold for fifty guineas by a souteneur, whose mistress and source
of hvehhood she has been for some years, to another souteneur whom
she hates, takes poison.
Act IV (A few weeks later). — A miserable hovel, in which Bamba's
uncle, recovered from his cocainomania, is looking after her. Her illness
has made rapid progress ; she is frightfully disfigured, and is only
able to crawl about the stage on all fours. Uncle and niece are to be
turned out of the house, because the landlord wants it for his son's
honeymoon and because the neighbours complain of the infectious
disease housed there. The sound of church bells is heard ; Bamba
complains that nobody ever taught her to pray. In any case, being a
-Muslim, she hates church bells, and would like to hear a muezzin.
Various members of the family from earlier acts appear, including the
Pasha, now repentant, his son the seducer of Act I, and his grandson,
Bamba's half-brother. The latter goes out to fetch flowers for Bamba.
Left alone with her uncle, Bamba asks him to put her out of her misery.
He consents ; she asks liim first to read her a few verses from the
Qur'an. Having done so he smothers her on the bed with a pillow, after
^'hich he himself goes raving mad.
Similar to this piece are " al Jahim ’, " Kukayin, ’ " Ibn as
Sifah,” “ Aulad a® Dhuwat,” Banat al Yaum,'’ " as Salib wal
998
X. BARBOUR
Hilal ” (a Copto-Muslim love story, forbidden by the censor). 1 i»'st
that can be said for these melodramas is that they fill the thc,i: ami
give it some much needed financial support ; moreover, tho^e a can
stand the succession of horrors will be rewarded by seein. am*
excellently acted scenes and some bizarre aspects of Eg}"ptiaii lin uch
as the “ Zar ” in the first act of “ Aulad al Fuqara' ” and the w ling
in “ al Jahim
Romantic and Historical Plays. — Two romantic and four hm a al
pieces were played during the season of 1932-3. The two mn nc
pieces were old favourites ; of the historical pieces, one wa i.-'t
produced in 1931. and three were new. The two old favourite' -re
Salah ad Din, a stock piece of the company of Munira al Main e va.
and “ al Badawijya ”. The latter piece, written by Ibrahim Ih. a;i.
was first produced in 1918 by the company of ‘Abdurrahman Ifi i'h'
by whom it is still occasionally played. The theme is the carry is off
of a Beduin girl by the Khalifa al Amir bi'ahkami-llah, ami ter
resistance ; this gives scope for the uttering of many e.x. o'd
sentiments, somewhat in the style of the Spanish playwright Lop'
Vega, without the poetry which lifts the latter into the realms of l
literature. In fact, “ al Badawi)ya ” is in the tradition of al ^1*' '
wal wafa ’ ” of Bairut of the eighteen-eighties.^ Of the historical pn
the first in date and in some ways the best is “ al ‘Abbasa T l^t H ’
ar Ra^Id ” by Mahmud Badawl.® This play deals with the fa'
the Banu Barmak. The author accepts the story of the marriag' "
Ja'far to the sister of the Khalifa, but gives it only a secondarp p
in the Khalifa’s decision to rid himself of his too powerful M azir. I '
play is straightforward and distinguished by a fair sense of the theati' .
The first act shows Ja‘far conspiring against the Kh alifa and relea>in-
the ‘Aliyyid prisoner, having rejected the counsel of his father \ ah\ '
Incidentally we are introduced to an entertaining Majlis of the Mazu
with the poets whom he patronizes. The second act takes place in 11“
private rooms of the Princess al ‘Abbasa. Ja‘far with difficuh,'
persuades her to accept his point of view. Noteworthy in this act i-
the pretty scene where Ja'far and the Princess play a game of chc."--
* Matba'at as Saqur, 1922 (classical). Ibrahim Ramzi is author of *’ al Hakm'
bi’amri-llah ”, ” Abtal Manjura,” and various other adaptations ; and also of aii
amusing sketch in the colloquial, “ Dukhul al Ilammam muA zai Khuruju,” produce^
in 1917 and printed in 1924, al Matba'at as Salafiyya.
® There also exist in print several romantic plays, apparently adaptations,
.Tamil al Bahri, e.g. “ Sajin al Qasr ”, “ Qatil Akhihi.” Matba’at az Zahra, Haifa, 19- ' •
“ Printed in Cairo, 1931 (classical).
THE ARABIC THEATRE IN EGYPT. Ill
999
The third act takes place in Harun ar Eashid's palace and shows
Harun’s suspicion of his minister increasing until he decides upon his
overthrow. The fourth act, again in the Princess's apartments, reveals
the failure of Ja‘far’s plan, the arrival of the Khalifa, the execution of
•Ja'far, and the pardoning of al ‘Abbasa herself and her little son. A
second historical piece, “ Shajarat ad Durr,” ^ by the same author,
was produced by the Eamses company in the season of 1932-3. The
author has followed the historical authorities closely and tells in an
unpretentious way the events of Shajarat ad Durr's life from the death
of her husband al MaUk as Salih to her assassination of Aibak in th(‘
Cairo citadel. Unfortunately a certain prosiness, which was apparent
already in “ al ‘Abbasa ’', is here more accentuated, so that the play
fails to grip at any point. Nevertheless, its theme makes the play
interesting to those who care for the history of medieval Egypt.
The remaining two pieces were amongst the four which shared the
second prize in the dramatic authorship competition of 1932. The first.
“ al Had!,'’ ® by ‘Abdullah ‘Afifl, concerns the reign of the j^allfa
of that name. The author, as was perhaps natural in a former pupil of
al Azhar, exalts the character of al HadI and represents him as
struggling against the efforts of his mother to distract him with dancing
girls and other worldly diversions in order that she may continue to
direct the affairs of State herself. Frustrated in her efforts, she causes
him to be strangled. The author has not yet a very developed sense of
the theatre ; the opening dialogue between two maidservants and the
wailing at the end are too long. But he too can tell a story in a straight-
forward way, while his language is virile and poetic. The play therefore
was interesting and fairly successful. The last piece ^ was by the young
poet ‘Adil al Ghadban, recently passed out of a Jesuit college.
Sympathizing probably with the nationalist aspirations of modern
Egypt, his thoughts naturally turned to a Pharaonic subject, the
e.xpulsion of the Hyksos Kings. The author prefaces his play with a
list of historical sources, but the difficulties of reconstructing the
atmosphere of a Uttle-known period were evidently too great for him.
Moreover, he appeared to have been influenced unhappily by the
classical French drama, so that his piece was little more than a series
of isolated declamations, of which the style was in itself good, but quite
insufficient to create a play.
' Printed in Cairo, 1933 (classical).
^ Matba‘at al Ma‘arif, Cairo, undated (classical).
^ Ahmus al Awwal au Tard ar Rii*at. al Matba'at al ‘A^riyya. t airo. 1933 (classical).
1000
X. BARBOUR —
Plays of Modern Life. — Several of these pla vs deal with the - • of
a girl of good family who is married to a man for whom she d: lot
care. In the first act the heroine is already miserable ; in the ud
she is desperate ; in the third she dies of consumption, bv poi- or
by setting herself on fire. Of this general type are " Fati ' ^
produced in 1931, by Mahmud Kamil ; Gharlzat al Mara . - pn ‘ d
in 1931 by ‘Abd al Qadir al ilazini, and " Samira " (1933). bv li ' id
Hafiz. A certain number, however, show greater imaginative [ r.
and are enlivened by interesting dialogue. Of these we ma\ ae
Qulub al Hawanim (19.33), by Muhammad Kh urshid. Thi> i- h'
story of a married couple who have each a lover ; as the plot deV’ o-
the husband and wife come to understand each other's point of \ •
In the end, the husband divorces his wife, so that he can inarrv tie . li
he loves ; while the wife s lover is very reluctantly persuaded to o'li
his mistress into his legal wife. The situations are possible "1
interesting, the dialogue entertaining anti the conclusion satisfac’
There also exist in print two plays of Ibrahim al Misri whicli il
deserve attention. The first is " al Ananiyya ’’ (produced in 11'- ’
This is the story of a wealthy and completely selfish Pasha v ’•
having divorced two previous wives, has now decided to man '
young girl in addition to his present wife. The latter, determine , ■
prevent the marriage, succeeds in bringing together her stepstm i
Pasha s son by a former wife) and the girl in question, and in c<nt- -
them to fall in love with one another. In the end the Pasha is indie '
to gi\e up his design, but not until everv* member of the family I '
.suffered. The play gives a striking picture of uncontrolled egoism
a family where the impact of M’estern civilization has destroyed th
sanctions of Islani without creating anything else to take their jilai '
The second is Nahwa an Nur ' .* This deals with the life of a youm:
journalist, Muhsin, who sacrifice.s all worldlv interests in order t"
guard his independence and socialistic ideals. He is forced to live m
poverty ; when a rich newspaper proprietor at last offers him a goo''
post, vith the necessary guarantees of independence, he discovers that
his suppo.sed benefactor is really his wife s lov'er, and that his voimgei
brother, vlioni he has brought up with endless difficulties, i.s another
aspirant to her favour. Having refused the offer and ordered
^ Matba a JariJat a? Sabah, Cairo, undated (colloquial).
* Matba ‘at as Siya.sa and Matba‘at as Sabah. Both Cairo, undated (classical).
^ Printed in al Adab al Hayy ”, Cairo, 1930 (classical).
Printed in “ al Fikr wal ‘Alam ”, Cairo, 19.3.3 (classical).
THE AEABIC THEATRE IN EGYPT. Ill
1001
wife and brother from his house, Muhsin is left alone on the stage, half
delirious. As at last he decides to struggle on at all costs towards
truth, towards freedom . . . towards the light ". a shot is heard, and
Muhsin discovers that his brother has committed suicide. The piece
is gloomy, but undoubtedly powerful ; unfortunately the theatrical
managers did not see their way to produce it, possibly because the play
could only appeal to the small minority of the Egyptian public who
would appreciate a play in the style of Ibsen or Strindberg. It was
pre.sumably on this account, and also perhaps because of the socialistic
views of the hero, that the play was not awarded a prize in the dramatic
authorship competition.
Another interesting play of family life is " adh Dhabaih “ ^
(produced in 1925) by the late Antun Yazbak. This piece, like " al
Auaniyya treats of the misfortunes which befall a son owing to the
■‘elf-will of his father. The principal character is a retired general of
the Egyptian Army, not unkindly but determined to have his own way.
As a young officer, he loved an Egyptian girl, then forsook her to marry
a European. The experiment has not been successful : the desire of his
wife to “■ be free " causes never-ending friction. M’hen the Pasha's
■‘on IS already a young man, the old general can stand this life no more,
abandons his European wife and returns to the love of his youth. This
experiment is as unsuccessful as the first ; the son finds the separation
of his parents unbearable, and in the end conunits suieiile. The play,
apart from one exaggerated coincidence, is well constructed ; much of
the dialogue is well written. There is an interesting pas.sage in which
the Pasha explains his grievances again.st his European wife. " Your
second person plural. " he savs, '' has driven me mad : meaning his
wife s habit of treating him as the embodiment of Ea.stern husbands
ni general, and saying on all ocea.sions : " You. you. you . . . you're
all the same.'' The piece is said to owe a good deal to one or more
European plays. The same writer is also author of an earlier play
called hisifa fil Bait ".
There remain to be considered the plays of modern life of the late
-^luhammad Taimur. These are three. ” al '.\sfur fil Qafas (1918),'^
Abd as Sattar Efendi " (1918),^ and ’ al Hawiya " (1921) ® ; and
‘tre all written in the colloquial language. " Al "Asfur fil Qafas, in
' Shirka Matbu'at al Qirtas, Cairo, undated (colloquial).
' Al Masrah al Mi?rl ’ (vol. iii. Mu'.dlai'at Muhaiumad Taiiuur), Cairo. 11141,
I’P- 1-2.54.
^ Hayatuna at Tamthiliyya ' (vol. ii, Mu'allafat Muliaimu.id Taimur). Cairo,
pp. 327-452.
1002
N'. BARBOUR
foui acts, treats of the difficulties which are created for the ni); ri,-
year-old Hasan Bey by his father's miserliness. In his unhappiie rhe
boy falls in love with the Syrian maidservant ; the affair is discn , red
and the girl dismissed. Hearing that she is about to hear him a ■ .lid,
Hasan decides to marry her, with the result that he too is disniis.sp. om
his father s house. In the last act, father and son are reconciled ti" leli
the intervention of an influential Pasha who, after giving H n *;
father a lesson concerning his conduct to his son, promises to o ’ mi
for him a Go\ernment post on which his heart' has for vears bei i i t.
This is a well-written and agreeable play ; the characters oi lie
adolescent Hasan and of his father the miser are well delineated. ' he
close of the second act is striking, when the Pasha's first thought .i ■ r
turning his son out of the house is to summon his steward and i 'er
him to reduce the daily supplies of meat and vegetables by the
spending amount. The only serious defect in this play is the improl p h'
coincidence by which the intervention of the influential Pash i"
brought about.
Abd as Sattar Efendi,” in four acts, deals with the marriag - "t
the daughter of a member of the small bourgeoisie. ‘Abd as Snr’ n'
Efendi is an amiable but weak character with an ignorant p i
tyrannous wife. He is employed in the Ministry of Waqfs on a si i 1
salary. His son, ‘Afifi aged twenty-three, is an idler who lives at -
father's expense and tyrannizes over the whole family. He Inn -
friend whom he believes to be an influential and well-to-do Bey. b '
who is really a swindler and parasite. The plot deals with the effm i '
aided by his mother, to marry his sister Jamila to thisfrieiK
in the belief that his friend will in return arrange a marriage for liii '
with a rich Bey's daughter ; and the counter efforts of ‘Abd as Sattfn
to prevent this marriage and to give her to another suitor.^ The sudden
inheritance of a fortune by the second suitor is improbable ainl
unnecessary ; in all other respects, however, the play is master!' •
From the opening to the final page the language is crisp and tie
dialogue sparkles with humour and satire. The arrangement of the pla'
is excellent and the delineation of almost all the characters is admirable
Abd as Sattar himself, perpetually overborne by wife and son, and
occasionally ineffectively amorous towards the maidservant, is unfor-
gettable. So is the servant herself, calculating, unscrupulous, and
mahcious. The worthless son, who describes his occupation a.<
enfi ** Lebonnard, but the treatment
entirely different so that there is no question of imitation.
THE ARABIC THEATRE IX EGYPT. Ill
1003
amateur of the theatre, and member of the Society for the Prevention
of Cruelty to Animals is another excellent study. The following
passage, in which the family have to pretend to take an interest
in the health of ‘Aflfl s pet dog, is a good example of the author's
humour.
(<_) A)i jl) . 4Sjlc Ojlc LjI , ^_x_.U
VI L-! All ojjLc L/*^ ^
J-aIc J (a^i V) . cUfi. ^ 1 jVl
Vi . j'_^_A5l . ji '? Jjs
<iA_f i (Ali-V) IAjI
i-i . 03 JLAil U V. 03 A^TV . ^ Vi aJL
oAlc ^i^i . <X-J (3 ^-9 l» ^’1
iJL^i
c
L AAji^
A) Alj 3 ! li" i3 ? i 1) A 1 AXfl^ . 3*^
l " ^ jVii ^1.4 . jllji AaC-
^ii Vi i ^L« *•• ’
Ava^ a) 3^ ^ ^ • A-a)
1004
X. BARBOUR
j ^ yii l;l tl^jU A.y: col Uc
CiLl. j
c« I )jl . ‘
4^1 j . aJ U (_^ c^'
I'tj «^a)I CU)‘I^ -dU
! L dLit
Lf^ L^—^y„ ^'. \y ^
• 0-: Cj^ >X.j' (5 ^
«>
<0 I ^ a ^1 4^1^ ^U
4^11 S . ^L. I b . J
J'P liT
0*^*1: (s^. J ^ys-il <j^ J L . y^‘
(JIa' j ii . jirdi
‘-iU. I'l , t>j\s. 4dUo*a; (^|^]l li . (_^5Cl.l UL (jS^^ •
I 0^0* ^ ^ 0^9-=^ jj-« (>-:^-^ L .
*“V. *^1; lj cr^o" ^
THE ARABIC THEATRE IX EGYPT. Ill
1005
b IV .
t_j_yl2]! «\b.l ^
jllLl A,_| (UaSl J Cjj^>) . ^
jl ^ cs^l ^ ja[^ . jlUi
(^iUil J ^4 j )
pj’U) J)l <.J
iS ci ^ • jb*Jl A-.t
f ^ cr^« ^ W A. • iS _r^. • t5_r^: •
) . ojLiJt
Cl— 1 *A>“1 (_^t-\l^ AJ^ 1 , . ac.
t — <^3^1 . jll-Jl -Uc
^J^\ «jLiJI V L
<3JU:dl <U J>/l ^\j jl^l’b (ibU) . jliJI Juc.
(,^y) j) ^_) . <\J Jul-i J ji .
^ al Masrah al 3/i*5rr, pp. 171-3.
1006
X. BARBOUR —
This play takes a high place amongst broad comedies ; if t). ; .okeii
anguage of Egypt ever becomes recognized as a language of lit ituie,
there can be little doubt that *■ ‘Abd as Sattar Efendi ' will uk as
Its first classic.
A1 Hawiya," in three acts, is a vivid sketch of the k nfall
and premature end of a wealthy young man who takes to ■ .unp
e plaj displays to the full the admirable sense of the theaf jon-
spicuous in all Muhammad Taimur s work, while the delineatio’' : the
progressive degeneration of the young man is master! v. It the
unremitting gloom of a painful subject will probabl y cause tli‘ litic
to p ace it below Abd as Sattar Efendi ” as an entertainment “ ugh
its composition is perhaps more perfect.
Dramas of the late Ahmad Shauqi.—lhe published dr, itie
compositions of the late ” Prince of the Poets " form a group by . . ni-
selves.i They consist of one prose piece Amirat al Andalus ’ md
five verse dramas. “ Masra‘ Kliyubatra,” * " Majniin La ' " '
Qambiz •• 5 (Cambyses), “ ‘All Bey al Kabir,“ » and “ ‘Anta . ' '
These pieces have all been produced on the stage. The best kno- of
^ Laila, is a version of the storv of the hopeles- >ve
of Qais for Laila. The piece when produced has the effectiven- of
a pageant of Arab life, rather than a play. It is true tli.. ni
the first three acts Laila is a maiden wooed by suitors, in the lb ’ th
act a married woman, while in the fifth act she” is dead. Nevertln : '•
tfiere has been no change in Qais's attitude or in her own ; them , i'
een no dev elopment of character. The merit of the plav consist' lu
e beauty of the verse, with its new presentation of old themes
Arab poetry, as m the lover's complaint on p. 17 of the Arabic text
(S
VI j.,Jl Itj
am, ®‘ahiuna by AlHliirralitnAn Ahmad as .Sa atl (Cairo, undated), sci r,.-
KhauV *** ’ " t’ath a] Andalii? by FuUhl al
^ Cairo, 1932.
^ Cairo, 1929.
» tr,in.slation of this plav made bv Mr. A. J. Arberiy
«as published in Cairo in 1933. ‘
^ Cairo, 1931.
• Cairo, 1932.
’’ Cairo, 1932.
THE ARABIC THEATRE IX E(;YPT. Ill
1007
i. ijL CmA)^
<-^j -5 ^
W*^=^ Cf *
l_j_jS Vj ~'_
Cr^- ^ t^"*
^.?r- " ^ -> Ct* ^ ^
;_$i)l 'iZ>y yy ^ 4^)1 Uc
(sUDl Laj\J j ,* 3 JJ J— ^
If we are to seek a comparison in English literature, the
parallel would perhaps be a poetic drama of Dr\den, such as -
for Love ”, with its melodious verse, its literary reminiscences, and its
artificiality. An example of the latter in “ Majnun Laila ” is the scene
at the opening of the second act, where the maidser\ant cannot n
the heart of the roasted .sheep to give to the lovelorn Qais. This w o e
episode seems to serve no other purpose than, to lead up to t e
<-A» ^
1008
X. BARBOUR —
The piece is frequently played both by Fatima Rushdi's comp.'
and in the Ramses Theatre. The production at the Ramses in the wii. -
of 1931 contained some interesting features. The repeated entry of
white figure of Qais by steps from the auditorium was very effectr
while the representation of the Jinn by white-robed figures with ' i
animal masks made impressive a scene which bad producing can ea-
make ridiculous. Much less happy was the representation of Laii '
spirit in the last act by a magic-lantern picture of the actress who h.
taken that part, cast upon a screen at the back of the stage.
The other published dramatic compositions of Shauqi Bey have ti
same general characteristics as Majnun Laila ” ; their subjects a‘
sufficiently indicated by their titles. “ Masra‘ Khyubatra ” is the mo
successful, though some of the speeches in it are of excessive lengto
“ Qambiz,” which deals with the Persian invasion of Egypt by th.o
king, was subjected to detailed criticism by the well-known Egj’pl^'^’
writer, al ‘Aqqad, in a specially published pamphlet.^ If the criticisi;
was to some extent justifiable, the harsh tone in which it was expressc-i
was surely to be regretted in the case of an old and honoured poet who-'
mastery of the classical tongue was certainly an asset to the Arabic -
speaking world.
The weekly paper arRisala, in its issue of 1st October, 1933, printed
an excerpt from one of two verse comedies (■' al Bakhlla ” and a.'
Sitt Huda ’ stated to have been found amongst the papers left by
the poet after his death. These fines included the following dialogue
between mistress and maid concerning the vegetable known as
Bamya ’ , which will serve to illustrate the poet's skill in an
unwonted field.
; ^ ^
• ^^■'***^
" Riwayat Qambiz fil Mizan. ‘Abbas Mahmud al ‘Aqqad, Cairo, 1931 (?).
THE ARABIC THEATRE IN' EGYPT.
Ill
1009
^ -fc*! 1 i|.
it j^iiiii
1 Ai.1 (_).:► I
y_jy\
The publication of these two plays has been announced for an
ily date.
Imaginative. — In the spring of the year of 1933 the literary world of
.ypt was surprised and delighted by the publication of a play entitled
.A.hl al Kahf written by Taufiq al Hakim, Wakil of the Damanhur
. iu -piet. This writer, who had spent some years in Paris, was hitherto
jDwn in literary circles only as the author of an operette, All
’■iba." produced some ten years ago, and a comedy entitled The
■iiumcipation of Woman It now appeared that he had been spending
> leisure for several years in literary work, of which the publication
■: .Vhl al Kahf was the firstfruits, to be followed shortly by that of
' less than six novels. Ahl al Kahf treats of the Quranic story
uerally identified w'ith that of the Sleepers of Ephesus. In Taufiq
! Hakim's play the Sleepers are three, the Court ^Ministers Tlashilinj a
lid Tlarnu^, and the shepherd Yamlikha ; with the latter s dog,
htmir, as a fourth. Having taken refuge in the cave from the
I 'Tsecution of Decianus, they fall asleep, and are miraculously
preserved for 330 years. When YamlilAa goes out to buy food, the
'leepers are discovered and taken before the king, whose ad^isers
recognize them as the saints whose reappearance had been foretold
irom ancient times. They are accordingly treated with the deference
due to saintly persons. In reality, having been very human people in
tlieir former life, they are still dominated in their new' life by the
interests which occupied them when they fell asleep. The shepherd sets
■'ut to find his sheep, Marnu^ to seek his wife and son, Waiiillnya to
carry on his courting of the king s daughter for by a stran,^e
coincidence the reigning king has a daughter who is the double of the
long since dead and sainted princess of the time of Decmnus, and who
has for that reason been called by her name, Prisca. Within a day t e
lirst two are disillusioned. The shepherd cannot bear the inquisitive
1010
N. BARBOUR —
crowds that follow him wherever he goes ; even his dog suffer' n
the curiosity of the dogs of Tarsus, where the scene is laid. Ma' ^
becomes desperate when he at last finds in a forgotten cemeteia if'
tombstone of his son, who died more than two centuries ago. ! -t
the shepherd, then Mamush. return to the cave. Onlv i
dazzled by his princess, stiU struggles against his increasing bewil i-
inent. Finally he, too, undeceived by his beloved, rejoins his frit; '
Once back in the cave they fall asleep again, to awake two months I. t
and discuss what they suppose to be their strange dream. But d'
overtekes the shepherd, and the touch of his new clothes as the nt
two cover his face, convinces them that this was no dream, bu •
reality. Shortly afterwards they too die. Following the Quranic st"'
the king then comes to build a church upon the site of the cave .
the cave is sealed the princess Prisca, drawn by some affinity to 1. ■
ghostly lover, slips in to die with the three saints. Though written
a play this piece has probably hardly sufficient action to succeed in i '
theatre. It breaks, however, entirely new ground in modern Aral’'
literature. The story is delicately conceived, imaginative, and tli •
language impregnated with a gentle satire which makes it ven
attractive to read.
APPENDICES
Conditions from 1933 to 1935
^ In the autumn of the year 1933 the companies of Yusuf Wahbi and
Fatima Eu.^di were dissolved, owing to financial difficulties. The
Ministry of Education then endeavoured to persuade the two companic.'
to unite, promising them the entire sum available in its budget for
theatrical purposes (£1,500), if they would carry on for what remained
of the theatrical season. A'usuf Wahbi and Fatima Rushdi refused to
participate, the former regarding the financial assistance as quite
inadequate.
Finally a company was formed with the title “ The Actors’ Union ’
(Ktihad al Mimhaththirm), which undertook to give performances for
the remainder of the season, in consideration of an immediate payment
of £400, considered as rent of the Alhambra Cinema, which was adapted
as a theatre, and the promise of further aid which amounted in the end
w Effendi Tulaimat was entrusted bv the Ministrv
with the artistic direction of this enterprise.
THE ARABIC THEATRE IX EGYPT. Ill
1011
Lnder these circumstances, seven new plays were produced, of
.\ liich the majority were in Classical Arabic, according to the wish of
'-he Ministry.
L nfortunately the attendance of the public was disappointing;
lii^ Mas no doubt m part due to the absence of the ‘‘ star ’ actor and
II tress, and in part also to the inaderpiacy of the adapted cinema for
ny elaborate production. The " Union ” dissolved at the end of its
'liort season, leaving Cairo without any Arabic theatre.
Subsequently a committee presided over by Hafiz 'Afifl Pasha,
iiiiuerly Egyptian Mini.ster in London. Mas appointed to advise
'U the circumstances of the Arabic Theatre. Its report has been
i'ublished in full in the Ahra)n ncMspaper of the "ifith and 27th March,
l and its recommendations may be siumuarized as folhnvs : —
(1) Creation of a Xational Theatrical Company.
(2) Annual grant of .£15.000.
(3) Renovation of the Opera House and its utilization for the
performances of the suggested company, from the beginning of
October to the beginning of January, and again from the end of
•March to the beginning of May in each year ; the intervening
months to be spent on tour.
(I) Creation of a Theatrical Academv.
(u) Despatch of theatrical missions to study in Europe.
(6) Translation of European plays into Arabic.
(0 Temporary engagement of a foreign theatrical expert.
11
Some Common Theatrkai. Ter.m.s
. al Fast
al Mumuththil
al Jnmhur, al Miilafarrujan
as Sala {la Sala)
al Banimr, pi. al BanCunr (Buiiinoire),
al Maqsura
al Ltlj. pi. al Aliraj (Logv)
Shibak at Tudhakir
* haracters of the Play A shkh ds ar Riwdya
Comedy . . ' . al Masldh
fostumes . . Maldbis ar Riwdya
Act
Actor
Hicbence
Auditorium
LoMer
Upper
Box Office
Vi-'L. VIII. PAKT 4.
1012 THE
ARABIC THEATRE IX EGYPT. Ill
Curtain
as Sitara, as Sitar
Footlight.s
Amcar al Hafa
Limelight .
. Arn^t an Nur
Make-up
al Mdkydj (Maquillage)
Melodrama
Drama ‘anlfa, Drdma fdji‘a
Play
ar Rncdya
Powder
al Budra (poudre)
Producer .
. al Mukhrii
Prompter .
al Mulaqqin
Rehearsal .
at Tajarraba, al Brufa (la prova)
Role
. ad Dur
Scene (in play)
al Mashhad
Scene (on stage)
al Manzar
Stage
. al MamtJial
Stage Manager .
Mudlr al Masrah
Stall
. Kursi mutntaz
Theatre
. al Masrah (popularly, and in older books
al Marsah)
Tragedv
. al Ma'sdh
Wig ' .
. al Bariika (Perritca)
\ote . — Tile writer's thanks are due to the many Egyptians who have
kindly supplied him with information concerning the Arabic Theatre,
in particular to Mine Fatima Rushdi, Mahmud Bey Taimur, Yusuf
Wahbi. Zaki Tulaimat, Muhammad Amin Hasuna, Hasan Lutfi al
Manfaluti, and Taufkj Habib.
“ Ju-sfaih Lun ” — a logical treatise ascribed
to Vasubandhu
By Boris Vassiliev
I
History of the Text axd Problems Connected with it
I CHINESE scholars have lately devoted considerable attention to
the history of Logic in their country, more especially to the history
'f its Buddhist branch.
Hsiien-tsang's school, which was translating and commenting on
le Buddhist Canon, drew its attention to the logical literature, issuing
• whole series of valuable translations and commentaries. It is to that
hool. so far as we know, that belongs the first attempt of the Chinese
" ^iiclerstand the history of Indian logic. Chinese tradition considers
■ hmiaga s system, partly completed by his disciple Sankaraswami,
"he the final stage in the development of Indian logic. Chinese authors
■ new nothing of its further development. Logic before Dignaga.
" cording to the Chinese conception of it. was divided into two periods :
he tir.st — non-Buddhist period — connected with the name of
^h.sapadad and the second — Buddhist one — connected with that of
' 'isubandhu.^
The name of Vasubandhu marks a new period of Buddhi.st literature.
1 he same must be said of his logical works, which are ascribed to him
China and preserved in Chinese translations. Concerning them we
'h'o gather information from such Chinese sources as occasional
'' ''fences, Cjuotations, etc. In general Vasubandhu's philo.sophic
' gacy m China con.sisted of thirtv-six translations of his works,®
treating the questions of ontology, cosmology, and dialectics, while
'a the realm of pure logic it is, unfortunately, insignificant,
Be learn that he had written three logical works : (1) Lunshih,
Zumu in Chine.se tran.slation.
P o-su-pan*ton in Cliine.^e tran'scTiption. Ticn-chin or Siiih-ohin in Chinese
■■'inslation ; ef. Vasubandhu's liiographv : HEFEO., vol. iv, 1904, p. 40.
tf. li.X.. App. i, pp. 371-2.
1014
B. VASSILIEV' —
(2) Lun-kwei, and (3) Lun-hsind The works themselves have 'v -
appeared ; we possess but fragments, scattered in scanty quotation-
Notwithstanding their complete disappearance, the ChiL* - ■
tradition has preserved another work which is ascribed to \ asubambi'
and which to a certain extent characterizes his logical concepts n -
That is the so-called Tarka-sastra or Ju-shih lun. It was mention i
for the first time in European sinology by St. Julien.- It figures uml' ■
the same title in the Catalogue of the Chinese translation of tL
Buddhist Tripitaka, compiled by Bunyiu Nanjio,^ who derives lu-
information from the catalogue of Buddhist books, Chih-yuan fa-pu >
k‘an-shu tsung-lu,'* composed in 1285-7. We read there : " In Sauskn’
it sounds Ta-la-ka sha-hsi-t‘e-la (Tarka-sastra). Ju-shih lun in thi'<
chapters. Chen-ti (Paramartha) of Leang translated. It agrees witl
Tibetan.”
The now existing te.xt of Ju-.shih lun “ appears in the index of tin
Chinese Tripitaka ® under two titles : Ju-shih lun and Ju-shih lun fan-
chih nan-p‘in. The second title is given as its complete name.
The text of the treatise is given in vol. xix, fasc. 1, and consists ot
12 pages with 10,536 characters.
Vasubandhu’s authorship is not mentioned, but in the footnote we
read that .some other editions mention it.
Contemporary Japanese reference-books on Buddhism contain
bibliographical information that is always based on the same old
Chinese catalogues. Thus in Bukkyo ta'i jiten — a Buddhist
cyclopedia — we find the following note : Nyo-jitsu ron (Ju-shih
' According to Tucci : Liin-shih = \ Lun-kwei = Vdda-vulhana, Lun-
lisin = Vnd(i-hrda)jn. But I’rof. .A. Vostrikov }ia> proved tliat tlie first is the 1 adati.
dhdtm and the .sei ond the ] rldandi. In EiirojH-an literature we have a detailed analysis
of the identification of these Chine.se titles in l’rofe.s.sor Tueci’s article “ Buddhi.-t
Logie before Diiinaga " (see JIIAS., July, 19-!), p. 48:1). He had already touched on
the subject in his article '■ The Vailavidhi ” (llIQ., 1928, v. iv, p. 63o). See also the
work of A. Vo.stnkov, Logical Work's of Vasubamlhu, eh. i. Mdy'abhusana in his
History of Irulian Logic, p. 207, tran.slates the third title, Lun-hsin, as Vada-kausaba.
We have a mention of these titles in C’liinese literature in Dignaga’s text : ^«yaya■
dvara (or mukha)- ” C'heng-li men lun ’’ and in the oommentaries to Nyiiya-pravesa-
"yin-ming ju eheng-li men lun ”, written by Hsiien-tsang’s disciples. Bee below.
^ St. Julien. Jouin. As., iv sene. vol. xiv, p. 3.39, Xo. 245. Jou-chi-lun. 1 livre
(Tarka-95stra) traduit sous les Liang par Paramartha.
® Bunyiu Xanjio. A catalogue of the Chinese translation of the Buddhist Tripitaka.
1883, London, No, 12.32. Zu-shih lun. “ Tarka-sastra.” Composed by the
Boddhisattva Vasubandhu. Translated by Paramartha, a.d. ;3.30, of the Khan dyna.sty,
A.n. 557-589 ; 1 fasciculus, 3 chapters. It agrees with Tibetan.
“ Cf. Appendix, No. 11.
Dai Nihon Kotei dai zo Kyo, vol. xix, bk, 5, pp. 67-73.
' p. 212.
" JU-SHIH T.rx
1015
.ii) — a book in one fasc., coniposed b\’ Boddhisattva Vasubandbu.
n ainartha of the Ch'en dynasty translated it. It explains the existence
r rao-Ii {yukti), the absence of it and the twenty-two '' Nigraha-
iiana's This note further points out the existence of a commentary
that treatise, written bv Paramartha. but the fact of that
aimontary having been lost is not mentioned.^
These are the only fragmentary data concerning the bibliography
' Thi.s text. Xo summarized de.scription of the data to be found in
il'erent Chinese catalogue.s has been made. A comparison of all the
' ts which can be picked up from variou.s catalogues of the Chinese
Cipintka and their arrangement in a chronological order will enable
- to form an idea of the history of the text of Ju-shih bin. since its
'st appearance on Chinese soil, and will allow ns to come to several
'iiclu.sions that will be of great use for the analysis of the content of the
■ xt in its present state. ^
According to the unanimous statements of all the Chinese
■ italogues, Paramartha ® translated the treatise Ju-shih Inn, and
■'i“refore it appeared in China during his lifetime in that country, i.e.
diring his activities as a translator, that is to say between A.i>. 548-569.
'Unyin Xanjio dates the translation at .r.n. 551).^
The same work was read bv a traveller. Dharmagupta. on his way
lorn India to China in 5£0, according to the statement of Professor H.
'■i in his Vaisesika Philosophy.'^
' We find similar .statements in tlie Huddliist evclopadia, Fo-h.'^io lu zu liiin.
edited in Shanghai,
■' t\’e find some data, unfortunatelv incomplete, concerning Ju-shih lim, and the
' orks connected with it in gralxxlh Chandra ll.agchi s J.e Vtnioii I!„ti,Wn(iuc m Chine
’ ‘ris. V. i. Sec Appendix 3 (f.n.)-
^ i^aramartha's stav in China eoineided " ith two periods of Chinese poUtital lustorA
"unected with the reign of tw(» dynasties. Leang and Chen. Ihus the period from
d.Stn .I,')? falls upon the i.eang and that from .mT to .169 ujion th en, therefore he is
' died .sometimes Paramartha of Cli'en. sometimes Paramartha of Leang, The Chih-
iiHii fa.pao kau-tung t.-nng-lu Catalogue asserts that the translation of Ju-shih lun
■ '■longs to the Leang period, when about ten works had been translated Iw Paramartha.
Il'>nyiu Xanjio bases Ins statements eonrermng chronology on the same Chinese
‘Ualogue. See Appendix. J^ee it.A'.. App. ii. pp- 4:13-3. About Paramartha s
■ ‘"i;raphy see BEFEO.. v. iv, 1904. pj). 3 and 60. and Bagehi, p. 41S.
^ ff. App. i, p. 372.
^ C’f. O.T.F.. No. 24. ri. Fhih>snphy. p. S4, f.n. 4. - Dharmagupta, a
Hiiddhist of Southern India, eaine to China in .i.i>. 390 and died in 619 («..\ ApP- n,
!'• 131). According to his life tJl.X.. Xo. 1.493, p. 92''. P-
"ork m Sha-16 (a province of Chinese Turkestan), when on the way to China. W hat he
read consisted of 2,000 slokas. But the extant work consists of about .130 s okas,
‘''orae of the oldest catalogue.s, Xos. 1,604, 1,609, mention that the book is m tw^o vols..
I'd the extant one is in one vol. And the beginning of the work clearly sho«s that it is
1016
B. VASSILIEV —
e learn from the K=ai-yuan shili-chia-lu catalogue and f
■ ch wan/ that Wing anv 1
^ e kmgdom of Sha-l6 (the present province of Hsin-chia i
harmagupta stayed, during two years, in one of the local tenn . ^
studjung among other books the treatise Ju-shih lun. consisting ■
-1.000 slokas.
Later on we have information of a merely bibliographical charac t
in lourteen dilferent catalogues.^
The earliest references belong to the Sui period
The catalogues ^Tos. 1, 2. 3, dated at the end of the sixth centin
and the beginning of the seventh a.d. refer to our treatise Ju-sln:
un as well as to three other treatises, namely Fan-chih lun {Pariprcchr.
/v ^ i^W^^f^^-sthdna-sdstra)^ and Cheng-shwo tao-li lu
yaya sastra) translated by the same Paramartha and closely
Ti, questions of Indian logic to judge after their title'
he length of each of the three is 1 fasc.^ All these treatises are lost.
1 necessary to mention in the first place the treatises To-fu hiu
and Cheng-shwo tao-li lun, which coexisted with our treatise Ju-shil.
un up o t e second half of the eighth century and then disappeared.
IS very probable that they were lost much earlier, because
Hsuen-tsang s disciples already knew nothing of them. One of these
i-h ^ ^ speaking about Vasubandhu’s logic, quotes
the second chapter of Ju-shih lun once, but he says nothing whatever
o e ree other treatises. He mentions the titles of Vasubandliu s
L . ^ ® ^ > Lun-kivei, and Lun-hsin, which had never existed
in a Chinese translation.
II to Hetuvidya-nyaya pravesa sastra.'^
as we as m his laisesika Philowphy,^ supposes that Cheng-shwo
aSremiritsf catdt ^
and (3> njgraha-sthVina Tlie'^l^ork '* refutation, (2) true refutation (of jitij,
doubtfui. Paramartha commented on uTumT Vasubandhu, but thi.s is
■He also translAtf'fl tLo v k i - commentan' (3 vols.) has been lost.
and the Cheng-shwo (or lunTtroti ™'-h the Pariprccha-sastra (1 vol.i,
the last.” ‘ ^ tao-ii-lun (I vol.), and he wrote a commentary (5 vols.;.
‘ Ci. Appendix. , t - ' t - ■
* Fasemdus = chUan " a roll ” m r ' ai^e.yila Philosophy, p. 84.
^ Kohuyaku dai JZo « -^ornpleted work,
pp. 52-6. ’ ■ ■*“ tan. Inmyo nyusho ri ron, Kaitai,
No
Hetnvidya ”, p, 55^ Ui’a article “ Introduction
1017
“ JU-SHIH LUX ”
ao-li lun could hav'e been a translation of Nyaya-sutra on account of
■IP fact that Paramartha wrote his commentary on it in five chapters,
■ liich correspond to five chapters of a commentary on Xyaya-sutra.
’p=;ide3, he beheves that Cheng-shwo tao-li is. in its turn, a translation
I the word “ Nyaya ”, as well as Cheng-li ”. Both these arguments
1 favour of Professor Ui’s hypothesis concerning the identity of the
■'t Cheng-shwo tao-li lun with Nyaya-sutra are not very convincing.
As to the treatise Fan-shih Inn, which had disappeared in the
■ihth century, together with the two preceding ones, it was merged
"ur text of Ju-shih lun, in Chinese catalogues, where it figures
'.der a new complex title, that of Ju-shih lun fan-chih nan-p in.
' hat was the connection between Fan-chih lun and To-fu lun. on the one
ind, and Ju-shih lun on the other, that is to say — have the two first
'pii irrecoverably lost or were they incorporated in some way into the
xt of Ju-shih lun and what might have been the original form of the
arka-sastra treatise is difficult to decide, since we have nothing but
ibliographical data.
Nevertheless, these data suggest the possibility of an incorporation
I the first two treatises in the latter, since Ju-shih lun in its present
tate also includes the elements of Fan-chih {Pariiceh'i) in its second
‘■'ipter, and of To-fu (Xigraha-stfiaDa) in its third chapter.
Passing to the text of Ju-shih lun itself, we see that at the end of the
xth century and at the beginning of the seventh this treatise figures
I'st in two fasc., then in one. with a special commentary on it, con-
'I'ting of three fasc., composed bv Paramartha.
At the end of the seventh centurv, according to the catalogue, s
'‘O'-. 4. 5, and G, Ju-shi lun is registered as a work in one fiisc. and in
-- nr 23 sheets. It is noteworthv that the above-mentioned com-
■ipiitary of Paramartha in three fa.sc. still existed at that time. But
""ards the second half of the eighth centurv or the beginning of the
nnth, the general situation undergoes a radical change. According to
lie catalogues Nos. 7. 8 and 9 Ju-shih lun in one fasc. remains the
nip preserveil treatise and is known under the comple.x title of Ju-shih
nm fan-chih nan-p'in.
The other three, as well as Paramartha s commentary, disappear,
diile they temporarilv remain in the list of lost texts with a note
xplaining that Fan-chih lun must have probably been the very Ju-shih
an nhich now exists in the Tripitaka. In the eleventh and twelfth
* '"'fiuries, according to the catalogue No. 10, these other treatises
'-a.se to be mentioned even among the lost works, and we have but
1018
B. VASSILIEV —
Ju-shih Inn in one fasr. This is evident from the followi
table :
^ Xos. of
1 cata-
i logues
The end of the 1
sixth and the
beginning of
the seventh
rentnries.
The end of the ' 4
seventh cen-
tury.
7 *Fan-chih lun in 1 fase. They now suppose that the j^esen
treatise is nothing but the veiy’ Ju-shih lun which is '*
be found in the Tripitaka. That is why its title runs .
" Ju-shih lun fan-chih nan-p'in. Paramartha of Ch en
translated. Separate volume.”
*To-fu bin in 1 fasc.
*Cheng-shwo tao-Ii bin in I fa^*.
H ... Deest . . .
The end of the b
eighth and
the beginn-
ing of the
ninth een-
turie.N.
Eleventh-
twelfth crii-
turie*-.
Id
. Dei .t . .
Thirteenth
1 1
. Dec-..f . .
century.
The end of the
fourteenth
and the be-
Id
. Deest . .
ginning of
the fifteenth
century.
*Fan-chih hm in 1 fasr. (The same note as in the catalogue
Xo. 7.)
*To-fu hin in 1 fasc.
*rheng shivo tao-li Inn in 1 fasc.
The middle of
the eighth
century.
Fan-chih lun in 1 fasc. Translated by Paramartha of Oh t :
To-fu lun m 1 fasc. Translated by Paramartha of Chvn
Cheng-shwo tao-li lun in 1 fasc. Translated by Paramarri
of Ch'en.
To-fu lun in 1 fasc.
Fan-chih lun in 1 fasc.
Cheng-shwo (or lun) tao-Ii Inn in 1 fasc.
. . . Dei st . . .
Fan-chih lun.
To-fu lun.
Cheng-shwo tao-li lun.
Fan-chih lun in 1 fasc.
To-fu lun in 1 fasc.
Cheng-shwo tao-li lun in 1 fasc.
Fan-chih lun in 1 fasc. Translated by Paramartha of Ch eu.
To-fu lun in 1 fasc. Translated by Paramartha of Cndi.
Cheng-shwo tao !i lun in 1 fasc. Translated by Paramar i i
of Ch en.
• They are marked in the list of the lost works.
■■ JU-SHIH LUX
1019
'll the catalogue Xo. 11 of the thirteenth century we find Ju-shih
tiigether with a mention of its Sanskrit equivalent, viz. Tarka-
■ ra.
riie catalogues Xos. 12. 13. and 14. in the fourteenth and fifteenth
: uries laconically inform us of a certain Ju-shih lun in one fasc..
anut any special comments or references to its translator,
'summarizing all the bibliographical data, concerning this treatise,
iicir chronological order we shall have the following table of its
uirical existence on Chinese soil : —
Jsos. of
cata-
logues
■' middle of 00 Ju-shih lun tran.'-lated by Paramartha.
'ho sixth
•■nturv.
end of the 0 Ju-shih lun in 2,000 .slokas existed and was read by
'ixth cen- Dharmagupta.
eirv
end of the
;i-xthandthe
'’Cginning of
'he seventh
' entunes.
1 Ju-shili lun in 2 fasc. Translated hv Paramartha of Ch'en.
Commentary on Ju-shih lun in 3 fasc.
2 Ju-shih lun in 1 fasc.
Ju-shih lun ill 2 fiisc.
3
Jiieendofthc 4
seventh cen-
tury.
Ju-shili lun in 1 fasc. (23 sheets). Translated by
Paramartha of Ch en.
Ju-shih lull in 1 fasc. Translated by Param.'irtha of Ch'en.
Conmentary on Ju-shih lun in 3 fasc.
Ju-shih lun in 23 sheets.
Ju-shih lun.
Ju-shih lull in 23 .sheets.
Ju-shih lun in 1 fasc,
Paramartha of Ch'en.
(22 shectf').
Translated
by
Ju-shih lun in 1 fasc.
Paramartha of Ch'en.
(22 sheets}.
Translated
by
li
ihe middle of
eighth cen-
TUry.
S
Ju-sbih lun in I fasc.
Ju-shih lun in 1 fasc. The title for it is : Ju-shih lun fan-
chih nan-p in. Pan-ehih lun. a co existing text, appears
in the enumeration of lost texts, figuring among them
with the following note : '■ They now suppose that the
Jiresent treatise is nothing but the very Ju-shih lun which
IS to he found in the Tnpitaka. That is why its title runs :
‘ Ju-shih hm fan chill nan-p in. Paramartha of Ch en
translated it. Separate volume. "
Ju-shih lull in 1 fase. The title is : Ju-shih lun fan-chih
nan ji'in m 2.3 sheets.
Ju-shih hm in 1 firse. Paramartha of Leang translated it.
1020
B. VASSILIEV —
Xos. of
cata-
logues
The end of the
eighth and
the beginn-
ing of the
ninth cen-
turj-.
9
Ju-shih lun in 1 fasc. The title is ; Ju-shih lun ; hii.
nan-p'in.
Ju-shih lun in 1 fasc. The title is ; Ju-shih Imi i iuli
HA-p in. Paramartha of Leang translated it. rdty
volume.
Ju-shih lun in 1 fasc. The title is ; Ju-shih fan-, h u
p*m (33 sheets).
Eleventh-
twelfth cen-
tury.
10
Ju-shih lun. The above-mentioned word " ju ' ■ m-
absence of incorrectness and shih ” means faultU-
Thirteenth cen-
tury.
11
Tarka-sistra Ju-shih lun in 1 fasc. and in 3 ■ •).
Oaramartha of Leang translated it. It agrees wit!
iibetan. ^
The end ol the
fourteenth
and the be-
ginning of
the fifteenth
centun*.
Id
Ju-shih lun in 1 fasc.
•
13-U
Ju-shih lun in I fasc. Translated by Paramartha of Cli . i
Ju-shih lun in 1 fasc.
Taking in consideration all the above-mentioned facts, tve come t..
the following problems with regard to the treatise Ju-shih Inn, know,
under the general title of Tarfca-sastra : first-what does the text o:
hi work represent, that is to say, is it really Ju-shih lun, a separate
iin 1 some other work, or several other works, united
t er IS itle, and third— supposing Tarka-sastra had existed a-
a separate treatise could not other texts have been inserted into
whaf '•ersion ? Assuming that its present form is incomplete.
that "iV°, are the parts
that were added to it later.
„f .11' *““!■' “»Jy «fte a thorough analvsi.
hV A P"'”"'' Tibetan tetu. But e'veu
' forth Ir'fT '»™al labhographical data enaWo ua to se.
thatifiia ‘l '-'P^ibotic atateiuouts. It goes n-ithout saving
0 lldil! 11 m i" 2-»00 slokas,
ird It fl . ITTT'” ' ‘“‘“'““y. «' the end of the sixth centurv.
.uiddll ol ,1 , ‘ ■‘PPtatance of its Chinese translation in the
the seventh e,’.”'' t" “bMtve that a century later toivards
eeventh c.n.^-y ,h,s .reatise exists in a very incomplete form.
“ JU-SHIH LUX "
1021
riie fact that ancient catalogues mention Ju-shih lun in two fasc.,
Ije considered as a confirmation of this hypothesis. Later on the
tise figures in one fasc. in 22 or 23 sheets. But it is doubtful that
the two original fasc. that figure in the catalogue should have
'i sonted a complete copy of our treatise, since, side by side with
111. we see three fasc. of the now absent Paramartha's commentary,
ii.sually to every fasc. of the main text there corresponds one fasc.
le commentary. Therefore the text must have originally consisted
iree fasc. with three respective fasc. of commentary. In its present
1 the text of the treatise is mixed, as is clearly shown by Prof. A.
-trikov, op. cit. It is perhaps blended together with the text of
-ehih lun, which had existed in the same time with our treatise,
' which had later on disappeared.
It is difficult to state exactlv whether To-fu lun (A igraha-sthd>ia
m) was a separate work or formed the material which was inserted
1 third chapter in the existing text of Ju-shih lun. But we have
ry reason to think so, if we take into consideration that the other
iti.se. Fan-chih lun (Pariprccha-sdstra),^ was declared already in the
ath century, soon after its disappearance, to be no other than the
-'hih lun in the form in which it was inserted into the Tripitaka.-
'■re ensues that Ju-shih lun was already included in the canon
a text very different from its original version of two centuries
'-■e.
ihis statement concerning the identity of Fan-chih lun with Ju-shih
! does not yet convey that we have the former instead of the latter,
f it hints at the possibility' of a blending of two. or even three, texts
' e add the treatise of To-fu lun.
Jho task before us is the localization and definition in the now
luting te.xt of Ju-shih lun of these inserted elements. But this is
Professor Ui defines its S.'inskrit equivalent in his bonk, I Philosoph ,
Here we note a reference of the cataloajue Xo. .■> (see .\ppendix, p. 2i>. f.n. 1 and -)
ilie seventh centurv concerning the d.iting of Par.imarth.i s translations of Ju-shib
• o'l one side, and Fan-shih hm and To-fu hin on the other, as this is connected with
' 'luestion of the dating of the translation of Tarka-Uistra into Chinese given by
^'1.1 iu Xanjio. The materials offered in this catalogue are derived from Paramartha s
'-nipliy, \ihich does not even mention our treati.-e. Hut the catalogue points out
Ju-shih lun and some other treati-ses were translated lietween .v.i). .‘idS-ood in
•‘■■ng-kwang-sso temple in the reign of the Emiieror Wii-ti of Leaiig. On his way
‘I k to India, after the completion of his work, Paramartha stopped in the above-
‘"'tiuned temple in 556 and lived then till .569. It is during this period that he
■uislated Fan-chih lun and To-fu lun.
1022
B. VASSILIEV
possible only through a study of Tibetan materials on earlv 1 lati
logic. I refer the reader to Prof. A. Vostrikov's work •• Logical a ik.
of \ asubandhu , which gives the solution of this problem.^
Thus if we accept the testimony of the existence of Ju-shih r. in
the time of Dharmagupta, i.e. at the end of the sixth century, we all
hate to admit a considerable change in this text and a po- .k
addition of new elements to it. especially if we take into consideric ' in
the fact that according to Chine.se sources in Dharmagupta 's tine it
consisted of 2,000 slokas. whereas in its present state it has but
slokas.-
But if we keep in mind that two of the three fa.sc. have been i 't
and that much Las been added to the preserved one. as it becoi '
e\ident from this analysis — we .see that the portion of the origi il
text in our treatise is but very .small. We may conclude from t .•
above-mentioned facts that even a superficial knowledge of the pni’ ■'
formal data, concerning the text, leads us to a .serie.s of hypotht - -
connected w ith its present state. According to our data on Ju-shih 1' i
and the two other treatises— To-fu lun and Fan-chih lun— the fir •
m Its original version wms a separate work as well as Fan-chih lun ai
To-fu lun, which were translated later.
It is quite possible that the two latter became nothing but sepanU'
chapters of some other work.
In its present version Ju-shih lun is first of all a mixed text, perhaji-
with the two above-mentioned texts ; secondly it is incomplete, an.
our mam ta.sk is to analyse its contents. Meanwhile, it become.^
necessar} to dwell on the general notions, concerning Yasubandhii'.'
logic, handed down to us by the tradition of Hsiien-tsang's school ol
philosophers and commentators in order to elucidate the dilferenr
materials referring to the history of our text.
■■ Logic before Ui,m5ga ”
loiriril « I liirka-sabtr,-! clenote.s it as a generic idea. a.s .‘.onie
T-irk4.<.l *>* tJie same time he considers that particulai
into .Sanskru ^ w' " i,”"! single te.xt -Khich he had fully transpose.!
-ji th K- ** he docs consider it to bo one whole, but does not ascribt
before fwi! «<■ it wa.s written hy some Buddhist author
analysis of it” ' t aw that an analysis of the history of the te.xt, as well J'
did not nav atT'' \ confirm t’rofessor Tucci's categorical .statement, wh.'
the preset philological materials of the catalogues, expounded
who dlpart^"lom (roisevla Philosophy, p. 8d, f.n. 41.
oi'e sloha. c o a amount of Chinese characters, taking 32 characters for
“ JU-SHIH LUX ”
1023
II
Hsukx-tsaxg's School ox Vasubaxdhu's Logic
Buddhist Chinese logic devotes its attention almost exclusively
a single period of its development in India or rather to a single
iiient of that development, namely to Dignaga's early work,
and Sankarasvami’s yuya-pravcsaP- Starting from
' unique basis Chinese Buddhist studies advanced in two main
. i ctions ; in the direction of translations of Indian works and that of
uinal Chinese commentaries.
The seventh century represents the highest point in the develop-
nt of Chinese Buddhist literature. That was the period of the
■'‘rary activity of Hsiien-t.sang and his school of translators. Only
nr out of the whole number of his adherents wrote works which are
' reotly connected with logic. Three of them are his direct pupils —
wei-chi. ^Ven-kwei, Shen-t'ai. One — Hwei-chao. is the pupil of
'-'wei-chi.^ The materials that are found in their commentaries on
'inkarasvanii's y yaya-pravesa are. as a matter of fact, the only basi.s
r all our knowledge concerning logic in China. They contain the
Mflition always referred to by contemporary Buddhologists. The
liinese Buddhist commentators, who wrote after Hsiien-tsang and
followers, when referring to logical problems, also t|Uote from these
■ lurces. All the materials of Hsiien-tsang’s school, on which all my
•tatements concerning Vasubandhu's logic are based, are but
flotations from the following Chinese works ; (1) Shen-fai s com-
■ I'Uitary on the Syaya-mukha. (2) Wen-kwei‘s commentary on the
Huya-praceia. (3) K’wei-chi's conuncntary on the S yilya-pi'dirm,
uul (4) Hwei-chao's two commentaries on the same work.^ Analysing
die data derived from the above-mentioned works with regard to
\ asubandhu s logic, we must admit that the.se data are but occasional
informations given in connection with an analy.sis of .separate passages
^ Cheng-li men lun-Xvayji niukha, aeeordiiig to Tueci. Cf. Ct. Tucci* The
'"ukha of Diynnga, 1930, Heidelberg. Otherwise Xyayadvara. H.X. 1223. 1224.
din-ming ju eheng-li lun, B.X. 1216.
^ K'wei-chi (Jap. Ki-ki) (632-082). C'f. Bukkyo jiten, p. 199 ; Ui's 1
^‘hdosophy, p, 1. f,n. 4, and Ui's On the Author of Mahiiyana i^utruhunkara, j). 220.
■''iien-t’ai {seventh century) ; cf. Bukkyo fiten, p. 624. ^^en'k\vei (seventh centur\),
Hwei-chao (? 714) ; cf. Ui's On the Author of Mahnyano sutrCAamUtra, p. 221.
* Shen-t'ai, " Jin-ming ju cheng-li-men lun shu chi ; ^\en-kwel, Jin-iuing
J’l cheng-li lun su ; K*wei-chi. ibid. : Supplement to TnpiUika, i, vol. lxxx\i, fa&c. 4.
Httei-ehao, - Jin-ming ju eheng-ii lun i tuaii : H%\ei-.-ha<.. " Jin-ming ju cheng-ii
Hm suan-yao ’* ; ibid., vol. Ixxxvi. fa.-ie.
102i
B. VASSILIEV —
in the w orks of Dignaga and Sankarasvami (that is — Nya-m-n
f Two passages in Dignaga’s Nyaya-muk
B-hich he mentions Lun-shih = Yada-vidhana, without namin
author, serve as one of the points of departure for our conclusion. ,
regard to 1 asubandhu s logic.
In the first sloka of Nyayn-muMa Dignaga says ; “ Thesi.
the other component parts of the syllogism form a logical dei,
stration The former, i.e. the thesis, is the object to he proved (1
cisputant) and not rejected by contradicting facts.” ^ He t
coiimient.s on the first words, declaring that this subject had I
treated m the work Lun-shih (Vada-vidhana) and others.^ In the ai
coi^emary on the last sloka of the treatise, speaking of the Nigrih
stfiana Dignaaa mentions again Yada-vidhana, saying that the questi ^
, ' in philosophy in the Lun-shih a
t +>> ^®ta that serve as a point of depart. ■
to he Chinese tradition. Shen-t‘ai, citing Dignaga s words from t ■
0 'a comments upon them in the following manner : “ Accord it
the probans (logical argument) and exam].'
, ^ ^ ® syllogism and are named proof or demonstration. L
u-/. i- preceding words he mentions Lun-shih and otb'
h JJ-’ ^ say— Lun-kwei (Yada-vidhi) and Lun-hsin (Yacl,.-
. a). hese three treatises were written by Yasubandhu.” "
Yi«! Hsiien-tsang’s school knew abou-
su laiidhu s logic and about his works that are now lost. But th.
ft. «a<I B.X. 1223. translated by
proof (mhann). Hcre'il’clM “
want to Drove in amn a ■ I. ^ ' po.siOon only that particular argument that «.
‘■""tW'JKton- (to it) cairexdude°('it)”''" argum"''
of a proposition q re propo.sition, etc. This means that through the formulation
.svIlocL T f That many terms represent the
’ Tucei tran.sla\os“'thL" m'"' ^is VSdavidhi, etc.”
'latU has been alreadv settld inih^ -'"reoier this erroneous formulation of the
Vadavidhi, etc.”, and adds that for 7'" - ei,"’" commentary upon the
Professor Tticci i.s rieht in hw r ■ translation is doubtful ”. I think
First of all the text ha- -7" translation of this passage is wrong.
N-viv»,,*h. ” r r""*" ■■
quite possible Thus Di i t racter shu the character “ tang ”, which
l^ted as ” associ!,e n ir ^ - “ P'-g shu ” cfn be tram-
'he word ■ wo " is no-se^ - Va.subandhu and not Digniga and
'n the Japanese article of ™' ' Tti^ same interpretation 1 have found
kl. «,“* S*~' “ " in v„,. V, p. «2, .nd p, K
• S-PPI. .0 ,,
lha
in
Oin
■ ith
ml
■n-
-n
' n
h
■II
1
JU-SHIH LUX
1025
£(
J J
' 'tion is whether they were acquainted with his works from first
- or they knew about them from hearsay. Hsiieii-tsang's disciples
■ iition Vasubandhu's following works : (1) Lun-shih. (2) Lun-kwei.
Lun-hsin, and besides them (4) Ju-shih lun. Nevertheless, we find
' 'it quotations only from Ju-shih lun and only from the second
pter of the now existing text, whereas of the three other treatises
■ the titles are mentioned as evidence. They are mentioned in
. :iection with the above passages from Dignaga. Does it not ensue
r they had in their possession nothing but Ju-shih lun (and it is
iiestion what was its form), and knew about the existence of the
t of Yasubandhu's logical works indirectly from hearsay ^ or from
stations to be found in some other works.
^Summarizing all the data about Yasubandhu's logic which we
' ive from Hsiaen-tsang's disciples, I want to point out that their
rest in Yasubandhu had never been independent, but was like
it they had for example in Aksapada, who is mentioned in the same
” as Yasubandhu, that is to say, only from the point of view of
Torical moments preceding the development of Dignaga’s logic.
These data refer to the following points —
(1) they bear witness of the existence of the treatises Lun-shih,
• n-kwei, Lun-hsin, and Ju-shih lun, and of the fact of their belonging
\ asubandhu’s pen ;
I-) they prove the e.xistence of the three-membered and five-
uibered syllogism that Yasubandhu operated with, and
(■1) at last, they offer some materials of a general character.
Y en-kwei, commenting on the statement of ^ yaya-pravesa about
■ existence of two kinds of example : the homogeneous one and the
erogeneous one, refers to the polemics between Dignaga and
'^ubandhu, and quotes Pramanasamuccaya, saying : Dignaga
’ll in Pramanasamuccaya: ‘In Lun-kwei (\ ada-vidhi) the jar.
■ knitted as substratum of the logical reason, is an example of
iiiiilarity. This treatise was not written by Yasubandhu or, at least
r had been written by him before he had become versed in logic.
I'tit when he had become so he composed Lun-shih (^ ada-vidhana).
' '^kis work the example of similarity is given in the words : YTiat-
^ ' ‘‘’ugiura in his Hindu Logic as preserved in China andJapan, p. .32, says : “ . . . when
' * '“'^'u-tsang was in India he saw tlirec books on logic ascribed to Scish (\ asubandhu ).
Ronki, Ronshiki, and Ronshin ” (f.n. 1, .Murakami's Immyojensho, 129,
'I'lna also speaks of this). If so, it must have been possible for Hsiieii-tsang to pass
‘ ‘^crtaiii information regarding these treatises to his pupils. Besides the}' must ha\c
St, own Rraraanasamuccaya, and could derive some facts from this latter.
1026
B. VASSILIEV —
e^er is created by an effort is non-eternal (non-perinanent). ' T. ^his
there is no contradiction in niv' opinion.' ^
So far w-e know this passage is not to be found in the Prami. a^v
muccaya as preserved in Tibet. A doubt arises whether Weii-kv • wa..
not repeating an incorrect cjuotation. Anyhow, in another passa ■ i>f
his commentary ■\^en-kwei e.xpresses again this point of view 'lat
Lun-kwei had been compo.sed when Vasubandhu was not yet a sell, i -
K‘wei-chi speaks of early Buddhist tradition in the following tei , v
“ Maitreya. Asanga. Vasubandhu had they not explained L jr.
when dealing with demonstration ? ' ' »
^ He also characterizes \ asubandhu s work in the following ten '
At first Aksapada defined truth and error, then Vasubandhu h-
pleted this in his works Lun-kwei and Lun-shih. In them he 1 dv
e.xpressed the general principles, but did not analyse them in detail '
This lack of system is further mentioned by K’wei-clii. who sa' ' •
Although I asubandhu in his treatises Lun-kwei. Lun-shih. etc. gl
all the rules fully, his .style is complicated and the meaning v
confused.” ^
Besides that K’wei-chi offers a direct testimony to the fact th i'
^ asubandhu already operated with the three-membered syllogi-!.'.
If in ancient times demomstration was supposed to consist of fo n
parts . thesis, reason, and two e.xamples — of similarity and of contra'i.
later on Boddhisattva Vasubandhu states in Lun-kwei and his otln ;
works that demonstration consists of three parts ; thesis, reason, an !
example. These parts are indispensable for that demonstration, ainl
form a sufficient basis for the object to be proved. Therefore h ■
mentions only three part.s. ^
At last K'wei-chi in his commentary takes up the above-meiitione.i
^yaya-»nll■ha. explaining Dignfiga's words concerning Lun-shih and
Vi; l/ '• Ixx.xvi, 4 , 1). Z-i-h. 0,1 the trm- relation betne.r
Wia.vnih, an<n..<(a-v„Iliana, of. l-rof. Vo.str.kov-s work quoteri above. Withregar.i
e.se «o reatisc.-i an interesting pa.s.s,-,ge from Fen-i ming-i chi (B.X. 164U|,
I should be mentioned, ft runs that fasubandhu.
are sZ' , ” I ‘ (Icduftion.s, wrote Lun-kwei and Lun-.shih whuli
Sankhva l*aramarthasa[)tati and were directed againat tfu
written d '"'n' Z ’ ^’^'■'""‘‘'■‘hasaiitati is Va.siibaiidhu s special work.
fusioro^f ? rf ® Siinkhva theorier The con-
tusion of facts in this Chinese version is evident.
- Suppl. to Tripitakn, i. vol. Ixxxvi, fuse. 4, p
^ Ibid., p. 35()a.
Ibid., p. 347a. tice al.so f.isc. .5, p. 426
I ^“PPl- to Tripitaka. i. fa.se. 4, p.'ssia.
> .. p. 349b. Cf. also p. 370a. Cf. also p. 3.-,0«.
“■ JU-SHIH lux '
1027
' rves that Lun-shih. as well as Lun-kwei. were eoinposed by
Mibandhu. and that logical demonstration was treated in
•md Thus wei-chi is the only one who pointed out the fact
\ asid^andhu s operating with the three-membered svllogism, while
' r commentators, such as Hwei-chao. his disciple, speaks of the
t niombered syllogism only.* Besides. Hwei-chao states that in the
■'(■rved works of Yasubandhu we have the five-n\end)ered con-
iUtion. Hwei-chao in most of his commentaries, referring to
I'ubandlm s logic, repeats the words that had been spoken before
Cl by Hshen-tsang's disciples — Kwei-chi, Shen-t'ai. and Wen-kwei,
lily quoting them.® Yet. with regard to the live-membered formula
'vllogism. he offers a series of separate statements ■* referring to the
' .itise of Ju-shih lun. which operates with the five-membered
• ll<>ui.sm. too." We find references to Ju-shih lun by Wen-kwei only,
' neither Shen-fai nor K'wei-chi ever mention it. He defines Ju-shih
c a.s a Buddhist metaphysical work, declaring that " the sutras
d siistras that had lieen translated before and are being translated
'' contain mainly adhyatma-vidya. i.e. Buddhist religious literature,
'icreas logic is a general science. Among those works is Ju-shih lun.
' ■ • • ® and points to Vasubandhu's authorship. " In the treatise
"i-shih lun, composed by Yasubandhu. the logical reason is said to
"'Ses.s three characteristics: Pak^a-tlfiarmala, Sap<d'':‘(i-t')jripf>. and
I pifksa-vijatireka. This treatise was translated by Paramartha in
'c Time of Leang. Analysing it we find that the contents of the treatise
' 'omewhat similar to what had been said by Dignaga with regard to
c Tiii'ee characteristics of the logical reason. The similar treatment is
-i' cn 111 Lun-shih (Yada-vidhana)." '
Tims We find here a literal quotation from the second chapter of
'c-shih lun in its present form.
besides, it is noteworthy that Wen-kwei compares Ju-shih lun to
hun-shih. and we see that Ju-.shih lun is being quoted, whereas Lun-
'"di is only mentioned. Does not there ensue that in the seventh
' titury, when the text of Ju-shih lun existed. Lun-shih was unknown ?
Ilwei-chao quotes Ju-shih lun as well when dwelling on the question of
^ ]). Sfo also j>,
“ Suppl. to Tri-pitnka, i. v. Ixxxvi, fast-, o, n. 404/f. ibnl . p. llUtn.
^ Cf. iSiippl. to Tripitaka, i, vol. Ixxxvi, ftisc. r», p. 404 h.
■* Ibid., p. 40^^fl ; ibid., p. 418f/.
Suppl. to Tnptfdka, i, vol. Ixxxvi, p- 40.>fj.
® Ibid., btsc. 4. p. :mft.
Suppl. to TripUakrt, i, vol. Ixxxvi, fast*. I. p.
'•'L. VIII. PART 4. bt)
1028
B. VASSILIEV
the wrong thesis, and declares that the fourteen kinds of the 1 ' i
(jKikmhhasa) may be compared to the theory of sixteen jati's gi\ > u
the second chapter of Ju-shih lun.i
Besides the above-mentioned materials, Hsiien-tsang's school, ii
persons of \\ en-kwei and Hwei-chao, also tried to define the exainpi 1
the Indian syllogism. According to Wen-kwei. Vasubandhu took he ■
example an object (jar), defined by its logical marks, i.e. considered
object as a substratum of its logical mark. Dignaga denied this p' r
of view,^ saying that in such a case we could easily come to absiirdi
because in this case it would appear that the object (jar) is a substrati i
of qualities, such as capacity to boil and to be visible. Then, if we co' 1
from the presence of one property of the substratum conclude to t
presence of its other qualities, and from the similarity in one respect >
a similarity in other respects, we would be able to conclude from t ■
impermanent character of sound to, say, its quality of being able : '
cook and to be visible, since the impermanent jar can cook and c. >
be %’isible.
Here we have in Wen-kwei's commentary two direct references t •
the question we are interested in : one concerning the works Lun-kwi
and Lun-shih. treated above,® the other concerning Ju-shih lun
namely treating the question of the explanation of the thrf
characteristics of logical reason.^ If we add to that some occasional
references made to Ju-shih lun by Hwei-chao, that are to be found it;
his conunentaries,® we shall have all the materials of Hsuen-tsaiig '
school with regard to Vasubandhu's logic. Thus we see that the
Chinese tradition about Vasubandhu s logic is reduced to the above-
mentioned materials. Summarizing them we must admit that they
are scanty and that their incontestability and value are questionable.
A thorough study of the materials of the Chinese tradition of tlie
^ Ibid., p. 4186.
- G. Tucci in his article, “ Buddhist Logic before Dinnaga,” p. 479, only
summarized the data given by K.''wei-chi and Shen-ti'ai when saying ” . . . gather
both from K wei chi (oh. iii) and Shen T'ai (eh. ii) that the theory ol the
was known to the ancient masters, who held two tliffercnt opinions about it, which were
not accepted by Jjinnaga. Some thought that the vipak^n is that which excludes the
sapaksa^ as well as the pak^y:!^ so in the syllogism * sound is non-etcrnal, because it i^
a product, like a pot. the npnk^a ether ’ excludes the contrary of the non-eternal
as well as of the pot. On the other hand other logicians said that the vipaksn is every-
thing except the ntm-eternal while for Dinnaga, as is known, lipak-^n i& yatra p<ik‘^o n't
vidyate
^ Cf. p. I.";, f.n. 1.
Suppl. to Tripitaka, vol. Ixxxvi, fase. 5, p. 3376.
i^uppl. to Tripitaka, vol. Ixxxvi, fase. o, pp. 423'j, 406«.
JU-SHIH LUX ”
1029
eiitli century convinces us that the Chinese did not so much operate
th textual data, as with the oral tradition or mere <piotations.
But as all the theories of modern Buddhologists with regard
' \ asubandhu s logic, in general, and his treatise Ju-shih lun. in
ii'ticular, are partly based on that tradition, for the fullness of the
'•ture we must analyse these theories and settle the question of how
i they have availed themselves of this tradition and what are the
'iclusions thev have made out of it.
Ill
I HE Questiox of Vasubaxdhu's Logic ix Sixological Liter.wure
At the beginning of the present article I have already spoke i of
he purely bibliographical data concerning Ju-shih lun, given first
,'■ St. Julien and then by Bunyiu Nanjio.
In 1900 Sugiura in his book, Hindu Logic ns Preserved in China and
iapand touched for the first time the question of a systematic sttidy
1 ^ asubandhu's logic, but based his analysis upon insufficient
materials. 2 He declared that ; ” When Hsiian-tsang was in India
snw three books on logic attributed to Seish (Vasubandhu). namely
^omki, Ronshiki, and Ronshin.” He made this statement in a foot-
note. ^ basing it upon Murakami’s Immyo-jensho, and added ; Dinna
dso speaks of this. "
Aot having Murakami's book near at hand I cannot verify it and
h'cide from what sources this extremely important information was
Icrived. But Sugiura's statement has been accepted in the Buddho-
iogical literature as an incontestable fact of Hsiien-tsang's testimony
‘O the existence of three treatises by Vasubandhu in India.* He
further states, according to K’wei-chi’s commentary, that Vasubandhu
maintained that a thesis can be proved by two propositions oidy,
Hud that therefore the necessary parts in a syllogistic inference are
'Hdy three.”
‘ p. 32.
Gt. Tucci in liis article Buddhist Logic before Dinnaga,” p. 451, points out
'flat the only source of Suali, Vidyabhusana, Keith, etc. was Sugiura, and characterizes
•“in as follows : “ But being himself absolutely without knowledge of orthodox nyayii
“id of Sanscrit, he is in his statements and in his translations very often mis-
leading . .
’ P. 129.
' We find the same statement in Hsie Meng’s book. Outline of Buddhism [Fo-hsio
lo katm, p. 33)_
1030
B. VASSILIEV —
At last he gives some materials coacerning Ju-shih lun. s.i} t
■' The only work that remains to us. from which we can learn anvf
of Seish's logic, is his polemic against heresies (Nyo-jitsu-ron). i
quotes the five-menibered formula of the syllogism, adding : >
must have been the form of reasoning used in debate in those d.i
and since in this book Seish was not concerned with theoretical k-
and since Hindu logic is primarily practical in its purpose, we can!
disprove the statement of Kwei-ki by citing this formula.
Thus we see that Sugiura touclies but slightly on \ asubainlli
treatise Ju-shih lun (Xyo-jitsu-ron). which, according to his own won.
he considered to })e a single text, directed against heretics, witln^
expounding its content with the exception of a mention of the hN
membered syllogism formula.
Meanwhile these superficial data are fully contained in Suali -
Keith’s, and Yidyabhusana's works, where Sugiura s words tlm
Hsiien-tsang, travelling in India, saw three treatises by Vasubaiidln
and that the latter knew two forms of syllogism — the two-memberc'
and five-membered forms are repeated — with the addition of th>
bibliographical data, derived from Bunyiu Nanjio.
G. Tucci, who dealt with the immediate Chinese text, remsed to
a certain extent the data presented by Sugiura. In his article, A
fragment from Dinnaga,” ^ he giv'es for the first time a Chinese
equivalent in characters for Ron-ki. Ron-shiki, and Ron-shin in then
Chinese transcription, i.e. Lun-shih. Lun-kwei, and Lun-hsin, and trans-
lates these into Sanskrit. But such an identification as already pointed
out by Prof. A. Vostrikov is wrong (namely Ronki is not Lun-shih but
Lun-kwei, and Ron-shiki is not Lun-kwei but Lun-shih).’ M e may
derive from this article that Tucci is familiar wuth the materials on
Vasubandhu given by Shen-t‘ai and K‘wei-chi. In his other article,
“ Vada-vidhi,” ^ he tries to analyse Tarka-sastra (Ju-shih lun), and
declares that it has nothing to do with Vasubandhu's Yada-vidhi.
At the same time he informs us that he has translated Ju-shih lun
into Sanskrit. Thus we gather that he probably considers this text
to be one single text, although we know' that even Ui questions this.
I have already spoken about the doubtful unity of the text. Tucci »
utterances are still more categorical in his article, ‘‘ Buddhist Logic
» JRAS., 1928, p. 383.
* Cf. also the Japanese article of Professor H. Ui, Iiulo tetsugaku keiihyu, vol. v,
p. 347.
’ I HQ., vol. ir, p. 636.
jr-SHIH LUX
1031
’."re Dmnaga ".i He is quite positive in his statement that the
-itise does not belong to \ asubandhu's pen. seemingly giving little
di to the testimonies of Hsiien-tsang's school, but at the same time
^akes partly into consideration the fact pointed out by Hsiien-tsang's
‘ "ol that \ asubandhu operated with the three-membered formula.
bases his argument upon the fact of Ya.subandhu's adoption of the
I ne-membered syllogism, while we have the five-membered formula
■bi-shih lun. This points out its earlier origin. We read on
-'e 483 ; ” Ye do not know its author, but it is evident that the
■^ent redaction of the text, as it has been handed down to us, was
ntten by some Buddhist’’ — and further on. page 485; "We do
r know anything about the author of this book or its age. but we
• 1 } presume that it was anterior to Dinniiga. It may be also that this
■ irka-sastra. or a redaction of it. was existent already in the time of
't.syayana. Tucci, to support his argument, compares a certain
issage from chapter 2 of Ju-shih lun to Vatsyayana's work, stating
leir absolute identity, and concludes : '• So we should be inclined to
mik that Vatsyayana and even the final redactor of the Nyaya-
i'tra knew, if not this same text, another of those Tarka-sastras
Inch seem to have existed long before Dinniiga and in which the
iticisin of arthapatti was already formulated. That we can speak of
irka-sastras and not of a single Tarka-.sa.stra is proved by two
' fcrences to them which can be found in the Pramanasamuccaya-
ftti. In both cases Dinniiga uses the plural. " Agreeing with Tucci’s
"int of view that Tarka-sastra is a generic name applied to any
'-ncal treatise — (that becomes evident from a mere accjuaintance
nil the Chinese catalogue Chih-vuan fa-pao k'an-shu tsung-lu. where
ic' same word Tarka-sastra figures in the titles of other logical works,
'nil as Nyaya-mukha. Xyava-pravesa. etc., therefore we have no
•'n^krit equivalent for Jii-shih lun and tlie title of it remains
‘inran.slated) — I, nevertheless, insist that the cjuestion of its author-
inp remains unsettled, for while there exist data, namely those put
"fth by Tucci, that speak in favour of his statement, there are others
bnf liave been partly mentioned by me. and that are partly quoted by
^ 1 ostrikov. In anv case Professor Tucci was the first among
'Jropean Buddhologists to give a most valuable material and to
■^press hypotheses that must be counted with.
\\ e have some more materials in European languages in the works
Takakusu and Ui.
1 JBAS., 1!)29. p. 4.->l.
1032
B. VASSILIEV —
Bat the most signiUcant of all is Professor H. Ui's Japanese a,
reat.ng \ asub.ndka', logic in his Introduction to Xvava-nr.s
f Japanese translation of the Buddhist Canon i
T, doubt in Vasubandhu-s authorshi;
Tii n order to prove this he giv'es a comparative stud'
cnnr r' T formula of Vasubau,
contained m the lost treatises, i.e. Lun-shih. Lun-kwei. and Lun-h-
he mformation concerning the latter he takes from the works
Hsuen-tsang s school. Taking the data olfered by this school for ,
In ■ T P' sa3-s ; "lasubandhu wrote three work.'
3 'In - d i.s stated bv Hsu,
sangs^ adherents-Wen-kwei. Shen-fai. and K-wei-ehi. ‘ Besi.
ese three vve find references to Liin-shih by DignSga, and this trear,
ust have had something to do with Dignaga's new logic.-
in Tiirli *^^.1 '“^“don the fact that Hsiien-tsang saw these treatr-
"l "" TO'drsdictorv statements of K ttei-' !'
alri .s' ’ probable already opran
L„1 I "’.'“-"“"'‘'"'''J .n<I of Wen-dvei. on .he otic
band who declares that it was Dimaga who was the first to establi-
loue, r'"™ "■« V.snba„dh„'s logic, em
Hvri ■ T™' "" “''e-ntembered .syllogism. L'i cite
n-kwe,., testtmony th., th. j.,, ,„„ds in L„„-hwei for sap.k,
1. „ T " ' '* "■’"-'Otoal and points out at the sane
me that, accord, „g to K'wei.ehi. .ksanga, and Vasubandhu tteatc
aadhana and d„.»„. as di/reren. tinngs « , -Rat this is ton illogie.
; j.».ya„yn,b,
bpJior,). l,rt»e<Ti the .,1,1 anrl tl.<- mn,' ^ .icford incr to Hsiien-t-saiii:
took ai] the five parts of thf* s ft ' ^ •'^nys that '■ Boddhisattva Asann-
-Viganiana, for y,'.' '’"'T' ^dlilharana, Upanaya. an.
the three part's of the sell '' "ork Lun-kwei also took for .Sidhan.i
and fdaharana ean he'L^^,,
'vrorig to take all the three or th a ’ I'ratijna i.s Sadhya. Thu.siti-
andSlK-n-t aitreatUintre -O
dilferenee hetaeen the old arnTthr'i ! H-siicn-t-sang's sehool saiv the
Miffiffentlv worked on Ho l-g.e_ and noted that the old logic niv not
this point S.ldh'ina ” h 'y' ^ U Kwei-ehi is wrong in hi.s treatment ot
to Dasana, in thil "L o first, .Sidhana as oppose,!
formula. But, if ue sneak ,/s' Ih ^ losOta'
figures as Pratnila an^ s-'lh' ‘“‘*’“"‘'1 "PF'-site of Sadhva. it is Sadhva that
VasuhhandureSe.ll,ihana“a: " therefore Asahga an,l
O'- five parts of the logical formula" TV and accepted all the three
took a.s 'Sadhana Hetii and opposes Sadhana to Sadhya, therefor,
Kwc-ch, made a confusion of ih referred Pratijna to Sadhya.
“ erpretation.” ’Ueas and therefore gave them a wrong
“ JU-SHIH LUN ”
1033
have been uttered by Yasubandhu liimsclf. " lie adds. Dwellino;
' Wen-kwei's quotations from Pranianasamuccaya Ui expresses his
iibts on Yasubandhu's authorship of Lun-knei. and emphasizes
' following fact: "'After having studied logic, he (Yasubandhu)
iiposed Lun-shih. where he asserts that a jar could not have served
homogeneous example (sapaksa). that, on the contrary, the sentence
'hatever is created by effort is non-eternal ' must have served as
h. Dignaga, in his work Pramanasamuccaya. says that
'ubandhu's argument is correct and agrees with his own opinion,
us we see that Yasubandhu paid attention to the logical .sense of
hiharana — major premise — and considered a thorough analysis of
■ery important. This fact marked the first phase in the change of the
'■-membered syllogism." Ui declares that it is not clear at present
lat A’asubandhu's logical system was. but if we compare Asahga s
'lilt of view with regard to the three characteristics of the logical
Ison 1 to AAsubandhu's point of view on the same subject, it will
'■ome evident that Yasubandhu ascribed a great importance to the
lalysis of these three characteristics, owing to which he inclined
'■''arils the three-membered svllogism. Thus we see how the idea of
lee characteristics of the logical reason (trirupa-lihga) ileveloped.
1 i further takes up Ju-shih lun. of whicli he says : " This treatise
- generally thought to have been written by Yasubandhu. and we have
'I'leuce that even Wen-kwei admitted that. Nevertheless this has
' \'er been proved. The now exi.sting treati.se is but a part of the
‘ liu work and contains onlv Fan-ehih nan-p in.- The first chapter,
liose beginning has been probablv lost, treats the questions of the
'istence or non-exi.stence of Yukti.^ The second chapter cla.ssifies
■ c fifteenth padartha ■* of Nyava-sutra. and points out the mi.stakes in
' objections of the opponent. Tlie third discusses Nigraha-sthana.-
'ud in its enumerations, definitions, etc., is identical to the sixteenth
'•idartha of Nyayasutra. giving different interpretations from the
"innientaries on Nvaya-sfitra. But even here we have many things
'iiat coincide.' ”
He further passes to points which in his opinion gave rise to the
I'lubts in A asubandhu s authorship of Ju-shih lun. Though Tarka-
' I'tra — he says — considers pratyaksa and aniimana to be important.
* Yin >an h.-iiaiiit.
= Fan-chih nan p'in.
Tao li.
^ Ti nei nan.
“ Ti fii ch‘u.
1034
B. VASSILIEV
nevertheless Vasubandhu in his work Fo-hsing lund altogether li-
the importance of all the four, i.e. Pratyaksa, Anuniana. Upatn,
and Aitihya. ”
Then Ju-shih liin. speaking of the three characteristics of the li
reason.- declares that it is right to take them into account in de!
and wrong to overlook them. It is very doubtfid. however, t'
Vasubandhu posse.s.sed so very exact a notion of these three elenie!-
I think he contradicts himself, saying that we must use the ti'
membered .syllogism if we mean to understand fully the logical val
of these three elements. It is most unlikelv that a man such
Vasubandhu. who considers that the basis of sapaksa is wliul
contained in the example, should admit such a contradiction. Anyhn'
the treatise recognizes three characteristics in the logical reason, hr.
on the other hand, u.ses the formula of the five-membered syllogisii
Besides in the catalogue of canonical books® this treatise does ii"'
figure as Vasubandhu's work. Moreover, the Corean edition of th
Tripitaka does not refer to it as Vasubandhu's work. Considering id
these facts, we have every reason to doubt that the given treatise wa-
written by Vasubandhu before we find new proofs of his authorship.
Thus we see that denying Vasubandhu’s authorship Ui at the same
time thinks the preserved text of the treatise to be one whole in it'
content, only incomplete in its form. Touching on Paramartha '
commentary. Ci points out the existence of such commentaries
in three fasc. to this work, which are now lost, and mentions three
other works — Fan-chih lun, To-fu lun, and Cheng shwo tao-li lun
he had translated, a.s well as a commentarv on it in fiv'e fasc. He say.s ■
‘‘ The question remains open as all these works have been lost, but we
have some ground to suppose that Cheng-shwo tao-li lun is a translation
of Nyaya-sutra. We may presume that Cheng-shwo tao-li is a Chinese
equivalent, as well as ' Chen-li ' to the Sanskrit word ' nyaya and the
5 fa.se. of the commentaries correspond to the respective five chapters of
the commentary to Nyaya-sutra.’'
Although it is a mere hypothesis, but with regard to Ju-shih lun.
the opening parts of it that have been lost, probably contained the
same data as the Nyaya-sutra. This hypothesis ensues even from the
presence of the theory of the five-membered syllogism and the four
kinds of yukti. An appreciation of Professor Ui's considerations is
‘ Fo hsing lun. li.N. 1220,
^ (a) Shih ken ben fa ; (6) t'ung-Iei so she ; (r) i-lei, hsiang li.
® Chu ching lu.
■■ JU-SHIH LUX
1035
■n by Prof. A. Vostrikov, ^ and I offer here a mere smiuiiarv of the
"iTial contained in Professor Ui's article so far as it refers to the
'tion we are interested in.
These are, roughly speaking, all the subjects that Ui touches on
onnection with Yasubandhu's logic.^
Thus we see that the Chinese tradition and all the considerations
iiodern Buddhologists that are based upon it. as a matter of fact,
very few objective materials for a detailed treatment of the
blem of Vasubandhu s logic in general and that of Ju-shih lun in
’.ticular, irrespectivelv front a rejection of Vasubandhu s author-
p or an admission of it.
Therefore without rejecting all the Chinese data and the conclusions
contemporary Buddhologists, it is absolutely neces.sary to solve this
' 'blent through a cross-examination of Tibetan materials and through
■uiaiysis of the text of Ju-shih lun it.self.
Appendix
‘'t of tfit‘ of th' Chinese Tripitakn trhen’ J x-sfnh hui, etc.,
arc minlinncd.^
I. Chung ching mu la.
Trip. V. .38, fasc. 1. B.X. No. 1609. Comp, in .\.D. -594. In
7 fasc.
See Bagchi. p. xlvi (2).
Fasc. 5, p. llln.
II. Li tai san pao chi.
Trip. V, 35. fasc! 6, B.N. 1504. Comp, in a.d. .597. In 15 fasc.
See Bagchi. p. xlvi (3).
' * t Jjirjiriil Workf of Vasuhaiulhu.
t have not lucntionefi two articles offinat importance, incliuled I)\ H.ti, in
'"I V of hi.s ,itudi€« of Hindu Phih-iophy. in .Japanese. One of these articles is
' "licated to ■' Pre-Digniiga's logic" and the other one to the analysis of • Xy.wa-
'cikha . Jteside.s we have a new work in Chinese dedicateil to the same question and
•mten by Professor Hsu Ti-shan (cf. Yeiiching, Joiirwul of Chine.^e Studies. No. 9,
‘■'•il). Thf Buddhist Liujical Treatises Madkyamika ami Yngacdra Schools before
‘"in'iijn." pp. 1828-1868 of this work .are sfiecially dedicated to \ asubandhu and his
■L'lcal works including .lu-shih lun, with re-edited Chinese text of our treatise.
‘ Prabodh Chandra Bagchi in Le Canon Bouddhiijtie nt Chine (Pans. 1927),
-li es a list of our treatise.s and the catalogues that contain them. Nevertheless, these
'• ‘ni, notwithstanding the general value of the work as such, are incomplete and do
■I'Jt give Us materials whence we could draw conclusions as to the historv of our texts,
rhus With regard to Ju-shih lun we read (p. 423) : Joii-cheii louen un chapitre
■ l73h, 6) ; NI, (7H,. ID) ; TK (Sob. l.D) ; KL (aSb. 9) dit le colophon de 1 ouvrage
1036
B. VASSILIEV
Fasc. 10, p. 656.
Fasc. 11, p. 766.
Fasc. 13, p. 92a.
III. Chung ching mu lu.
Trip. V. 38, fasc. 2, B.N. 1603. Comp, in a . t >. 603. In 5 fasc.
See Bagchi. xlvi-xlvii (4).
Fasc. 1. p. oa
Fasc. 1, p. 10a.
IV. Ta t'ang nei tien lu.
Trip. V. 33. fasc. 2, B.N. 1433. Printed in a.d. 664.
See Bagchi, xlvii (5).
Fasc. 5, p. 77a.
Fasc. 6, p. 93a.
Fasc. 8. p. 101a.
Fa,sc. 9, p. 108a.
V. Ku chin i ching Fu chi.
Trip. V. .38. fa.sc. 3. B.N. 1487. Comp, in 664-5. In 4 fasc.
See Bagchi. .xlviii (5).
Fasc. 4. p. S56.
3T. Ta chow k'an ting chung ching mu lu.
Trip. V. 38. fa.sc. 3^ B.X. 1610. Comp, in a.d. 695. In 15 fasc.
See Bagchi. xlix (.5).
Fa.sc. 6, p. 24a.
Fa.sc. 6. p. 2ija.
Fasc. 13. p. 61a.
A’ll. K'ai yuen shih chiao lu.
Trip. V. 38. fasc. 4-.5. B.X. 1458. Comp, in a.d. 730. In
20 fa.sc.
Sec Bagchi. xlix (10).
Fa.sc. 7. p. 60a. Ola.
pnrti- le litre : .fon-cheu Imien f.in tehe nan p in. L’oiivrage c.xiste. Xanjio l-
Tarka-g,l..itra. " We see that I’rahoilh C'h. Ha«ehi in his enumeration niention.^ o
respective i atalopue.s Nos. ' 2 . 4, .">. anil 7 — the rest are not mentioned. The same m .
be s.iicl with reearil to the three other works (p. 429) : " Ian tehen louen, en
ehajntre. LK (6.5//, 8) . XL (77i) ; TK (856) ; KL (61«, 1) fait remarquer qu il n
pas certain si I'ouvrage forme une partie du Jou chen louen.” ” lo feu louen, ei
chapitre. LK (6.5//, 8) ; XL (77/,) ; TK (S-Vj) ; KL (61«, 1).” - Tcheng ehouo tao
louen, en un ehapitre. LK (6.5//, 9) ; XL (776) ; TK (86a) ; KL (61a, 1)- The
mentioned catalogues are nothing but our respective catalogues Xos. 2. 4. o,
the others are not mentioned hv Prabodh Ch. B,igehi.
“ Jtr-SHIH LUX
1037
Fasc. 14, p. 9a.
Ease. 19. p. 486.
Fasc. 19. p. 676.
VIII. K'ai yuen shih chiao lu Ho ch u.
Trip. V, 38. fasc. 5. B.X. 1486. In 4 fasc. See Bagchi. p. 1 (11).
It forms the last part of the prcceiling catalogue.
Fasc. 2, p. 84«.
IX. Chen yuen hsin ting shih chiao mu lu.
Trip. V, 38. fasc. 5-7. Comp, in .\.d. 794-5. In 30 fa.se.
See Bagchi, 1 (12). It contains the translations made during
A.D. 730-794.
Fasc. 9, p. 49a.
Fasc. 10, p. 556.
Fasc. 22. p. 39a.
Fasc. 24, p. 566.
Fasc. 30. p. 956.
X. Ta tsang cheng chiao fa pao piao mu.
Trip. V, 38. fasc. 8. B.X. 1611. Comp, in lli»5 and edited in
1306. In 10 fasc. See Bagchi. li (15).
Fasc. 6, p. 25a.
XI. Chih yuen fa pao ken fung tsung lu.
Trip. y. 38. fasc. 8. B.X. 1612. Comp, in 1285-7. In 10 fasc.
See Bagchi, li (16).
Fasc. 9. p. 75a.
XII. Ta ming ch ung k'an .san tsang chtuig chiao mu lu.
Trip. V, 38. fasc. 8. B.X. 1662. Printed in 1403-1424. It is
but the catalogue of Tripitaka. eilited in 1368-1398. In
4 fasc.
Page 1166.
XIII. Ta tsang mu lu.
Trip. y. 38. fasc. 8. In 3 fase.
Fasc. 2. p. 87a.
XIV. Ta p u ning .sseu ta t.sang ching mu lu.
Trip. y. 38. fasc. 8. In 4 fa.se.
Page 101a.
A New Classification of the Constituents of
Spoken Japanese
By S. Yoshitake
TT is now a matter of common experience that it is often found
impossible to describe and classify linguistic material within the
framework of the classical categories of morphology. The most
rea.sonable alternative and one which has the merit of being objective
and practical is the strictly formal and positional technique which has
for many years been advocated by J. R. Firth, Senior Lecturer in
Phonetics and Linguistics. University College. London. He states that
in grammatical studies " mental structures must be abandoned and
research directed towards finding purely formal and positional
characteristics of the facts them.selves which, being differentiae, will
serve as a means of description, and as criteria for a congruent
classification ". And secondly. " he continues, " beyond this technique
of formal description, function or meaning can then be studied in
context on sociological lines, unobscured by categories serving any
other purpose." ^
The classification proposed in the following pages is based on the
technique mentioned above, and is intended to apply to that particular
Japanese dialect which is now spoken by those citizens of Tokyo
whose speech would appear to the majority of educated Japanese
as entirelv free from unusual features.- In order to evaluate the
proposed classification with its new set of discriminative terms it is
necessary to compare it with the conventional classifications adopted
by native and foreign grammarians. Although no two foreign or native
writers on Japanese grammar seem to agree in terminology, it would be
quite unnecessary to consider here all the varied nomenclatures used
bv them. I have therefore chosen as specimens the classifications found
in McGovern's Colloquial Japanese and in Yamada's Nihon Kwjoho
Kogi. the former because of its comparative simplicity and the latter
by reason of its popularity in Japan.
^ J. R. Firth’s review of H. »Straumann’s SewsjHtper Headlines, in En/jlish Sfndit':,
xvii, June, 1935, p. 112.
- This is a modification of the negative detinition given to Standard English hy
Firth in his most stimulating and instructive little book, Speech, p. 63,
]040
S. YOSHITAKE —
AVitli a view to faeilitatinji comparison I sliall divide the poe I
i la^sitication into two parts, broad and narrow. Inasmuch 't
the categories contain three or more groups or forms distinsnii-
oiie from another, it is convenient in the prc.sent di.scussion to sej
the minor divisions from the main .sections. In so doing redupln n
of some of the propo,sed nomenclatures is unavoidable, but in n
cases the conventional English and Japanese terms will be given iy
m the broad classification.
A. Broad CnASSiFiCATiox
Pro poml T 01)11 uiologt/.
I. Invariables
(1) Connect ibles
(2j Supplements
il. lariahles
(1) Part Variabl IN
(«) a-Yariables
{h) zeroA’anablc.s
Coiive»tio)uil So))ie)>clntiii ' '
Xouns, Pronouns, Numerals. A
tives, Adverbs.
laifien (Substance words), t ^
(Adjunctive words).
Case Particles, Postpositioji'
.sponsives, Interjections. Nn
Classifiers, Suffixes.
Joshi (Auxiliary words). >
(Affixes), Settoji (Prefixes). >
b)ji (Suffixes).
Yof/r)) (Inflective wonls).
Coasonant Verbs.
Yotlan Kalsiujo Dd^sht (Quadric '■
inflective variable words). A",
•s/a’ (Words of existence).
Vowel Verbs, The Pa.s.sive lotiii I
Potential Form, The ( aii'H
Form.
Kfitoi Ichidru) Kiitsiojo Do-d)) (i i'i'
unigrade inflective varial>le
S/ii)tiO Irhid(t» Iuits)iijo
(Lower unigrade inflective vane*
words). Fuk))gobi (Formati^e > u
ings).
NEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1041
Proposed Terminology,
(c) ku-Yariables
(d) Shortened Variables
(2) Total Variables
Con vcntional iS omen clat u res .
Real Adjectives : Simple Adjectives,
Compound Adjectives.
Keiyoshi (Modificatory words),
Fukugobi.
Auxiliary Verbs, The Auxiliary
Suffix.
Yodan Katsuyo Doshi, Setsuinei
Sonzaishi (Explanatory words of
existence), Keigo (Honorific).
Irregular Verbs.
Kagyo Sandan Katsuyo Doshi (The
la-hne trigrade inflective variable
word), Sagyo Sandan Katsuyo
Doshi (The sa-line trigrade in-
flective variable word), Fukugobi.
B. Narrow Classification
(With
Proposed Terminology.
I. Connectibles.
(1) qa-no-Connectibles
inu (dog), hombako (book-
case), ore (I), kore (this),
dore (which), hitotsu
(one), ni3iu (twenty), asu
(to-inorrow), hooboo (aU
directions), byooki (ill-
ness).
(2) ga-na-Connectibles
baka (foolishness, fool),
bikko (lameness, cripple),
otemba (romping, minx).
(3) no-Connectibles
tjotto (a httle), sukoji (a
Httle), Jibaraku (awhile).
(4) na-Connectibles
d30obu (soundness), taiben
(seriousness, very).
examples)
Convent ion al Nomenclat u res .
Simple Nouns, Compound Nouns,
Personal Pronouns, Demonstrative
Pronouns, Interrogative Pronouns,
Japanese Numerals, Chinese
Numerals, Adverbs.
Meishi (Nominal words) Daimeiski
(Substitutes for Nominal words).
Sushi (Numeral words), Fukushi.
Simple Nouns.
Meishi, Jotai no Fukushi (Ad-
junctive words of state).
Simple Adverbs.
Tetdo no Fukushi (Adjunctive words
of degree).
Simple Nouns, Simple Adverbs.
Jotai no Fukushi, Teido no Fukushi.
1042
S. yOSHITAKE
Proposed Tenninoh[iy.
(5) Direct Connectibles
kono (this), sono (that), ano
(that . . . over there),
done (which), moo (al-
ready). motto (more),
tfoodo (just), nakanaka
(certainly). pikapika
(“littering).
II. Supplements.
(1) Initials
hai, ie, aa, o-, go-, dai-.
(2) Finals
yo, ro, tomo, -san, -tatji,
-nin, -hiki, -hon.
(3) Movables
wa, mo, ga, no, o, ni, e, de,
yori, kara, made, ka, nee,
keredo.
III. a-Variables.
(1) The i-form .
kaji (lend), kaki (write).
Jini (die), yomi (read),
yobi (call), kagi (smell),
utji (strike), uri (sell), iki
(go), ii (say).
Con ventional Nonienclalures.
Demonstrative Adjectives. Interro
gative Adjectiv'es ; Real Adverbs :
Simple Adverbs, /o-Adverb.s. Ono-
matopoeic Ad\erbs.
Daimeishi no Rentaikaht (The forms
of Dnimeishi to be followed by a
, Substance word). Yd;/en noRenhu-
kakit (The forms of Yoyen to 1 h
followed by a Substance word],
Jotai no Fnkushl. Teido no
Fukushi, Chinjutsu no Fnkushi
(Adjunctive words of declaration).
Responsives. Interjections.
Fenkd no Fnkushi (Introductory ad-
junctive words). Settoji.
Interjections. Numeral Classifiers,
Suffixes.
Shujoshi (Final auxiliary words),
Kantojoshi (Interjectional aux-
iliary words), Setsuji. Selsubiji-
Case Particles, Postpositions, Inter-
jections.
Kaknjoshi (Auxiliary words of quali-
fication). Fnknjoshi (Adjunctive
auxiliary words), Kakarijoshi
(Nexus auxiliary words), Setsn-
zoknjoshi (Conjunctive auxiliary
words), Setsuzoku no Fukitshi
(Adjunctive words of link).
The Second Ba.se. Verbal Nouns.
Renyokei (The form to be followed by
an Inflective word), Juntaiyeu
(Quasi-substance-words).
XEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1043
Proposed Terniinoloqy. Conventional yompnclaturei.
(2) The e-form . . .
kase, kake, /ine, yome, Meireikei (The Imperative form),
yobe, kaqe, ute, ure, ike,
ie.
(3) Coincides witli (2) . The Fourth Base.
Jokenkei (The Conditional form).
(4) The a-form . . . The First Base.
kasa, kaka, Jina, yoma, Mizenkei (The ' not-yet-so "-form),
yoba, kai)a, uta, ura,
ika, iwa.
(5) The oo-form . . The Fifth Ba.se.
kasoo, kakoo, /inoo
yomoo, yoboo, ka^oo,
utoo, uroo, ikoo, iwoo.
(6) The u-form . . . The Third Base.
kasu, kaku, Jinu, yomu, Skushikei (The Final form). Ren-
yobu, kaqu, utsu, uru, taikei (The form to be followed by
iku, iu. a Substance word). Jantaiqen.
(7) The te(de)-form . Gerund.
kajite, kaite, Jinde, yonde, yonde, kaide, utte, utte. itte, itte.
(8) The ta(da)-form . The Past Tense Form.
kajita, kaita, Jinda, yonda, yonda, kaida, utta, utta, itta. itta.
(9) The tari(dari)-form . The Alternative (Frecpientative)
Form.
kajitari. kaitari, Jindari, yondari, yondari, kaidari, uttari, uttari,
ittari, ittari.
(10) The tara(dara)-form .
kajitara, kaitara, Jindara, yondara, yondara. kaidara. uttara,
uttara, ittara, ittara.
(11) The taroo(daroo)-forni — -
kajitaroo, kaitaroo, Jindaroo, yondaroo, yondaroo, kaidaroo,
uttaroo, uttaroo. ittaroo, ittaroo.
viii.. VIII. lARr 4.
1044
S. YOSHITAKE —
Proposed Terminology.
(2) Coincides with (1)
(3) The re-form
mire, dere, -rere, -sere.
(4) Coincides with (1)
(5) The yoo-form
miyoo, deyoo, -reyoo,
-seyoo.
(6) The ru-form
miru, deru, -reru, -seru.
(7) The te-form
mite, dete, -rete, -sete.
(8) The ta-form
mita, deta, -reta, -seta.
(9) The tari-form
mitari, detari, -retari,
-setari.
Convent ional N omenclat u res.
Meireikei.
The Fourth Base.
Jdkenkei.
The First Base, The Fifth Base.
Mizenkei.
The Third Base.
Shuskikei, Rentaikei, Juntaigen.
Genmd.
The Past Tense Form.
The Alternative (Frequentative)
Form.
(10) The tara-form . . —
mitara, detara, -retara, -setara.
(11) The taroo-form . .
mitaroo, detaroo, -retaroo, -setaroo.
V. ku-^'ariables.
(1) The zero-form . . The Stem.
ooki (large), haya (early), Gokan (Stem), Jotai no Fukushi.
omo (heavy), sama
(cold), -na, -ta.
(2) The i-forni . . . The f-form.
ookii, hayai, omoi, samui, Shuskikei, Rentaikei, Juntaigen.
-nai, -tai.
(3) The kere-form . .
ookikere, hayakere, omo- Jdkenkei.
kere, samukere, -nakere,
-takere.
(4) The karoo-form . .
ookikaroo, hayakaroo, omokaroo, samukaroo, -nakaroo, -takaroo.
NEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1045
Proposed Terminology.
(5) The kix-forni
ookiku, hayaku, omoku,
samuku, -naku, -taku.
(6) The o(u)-forni
ookiu, hayoo, omoo,
samuu, — , -too.
Con venlional iSomi-nclatHres.
The ^•(<-form.
Eenyokei, Mizenkei.
Renyokei.
(7) The kute-form . . Gerund.
ookikute, hayakute, omokute, samukute, -nakute, -takute.
(8) The katta-forni
ookikatta, hayakatta, omokatta, samukatta, -nakatta, -takatta.
(9) The kattari-form . .
ookikattari, hayakattari, omokattari, samukattari, -nakattari,
-takattari.
(10) The kattara-forni
ookikattara, hayakattara, omokattara, samukattara, -nakattara,
-takattara.
(11) The kattaroo-forni
ookikattaroo, hayakattaroo, omokattaroo, samukattaroo, na-
kattaroo, -takattaroo.
VI. Shortened Variables.
(1) The i'form
— , nari, — .
(2) The i-e-counterpart form
nasal, — , -maji.
(3) The re-counterpart form
nasare, nare, -masure.
(4) The a-form .
nasara, nara, — .
(5) The a-e-counterpart form
— , — , -mase.
(6) The oo-counterpart form
nasaroo, — , -majoo.
(7) The u-counterpart form
nasaru, na, -masu.
(8) The te-form
nasutte, — , -majite.
The Second Base.
Renyokei.
The Second Base.
Renyokei, Meireikei.
The Fourth Base.
Jokenkei.
The First Base.
Mizenkei.
The First Base.
Mizenkei, Meireikei.
The Fifth Base.
The Third Base.
Skmhikei, Reniaikei.
Gerund.
1046
S. YOSHITAKE —
Proposed Terminology. Conventional Nomenclatures.
(9) The ta-form . . The Past Tense Form,
nasutta, — . -ma/ita.
(10) The tari-form . . The Alternativ'e (Frequentative)
Form.
nasuttari, — , -majitari.
(11) The tara-form . .
nasuttara, — , -ma/itara.
(12) The taroo-form . .
nasuttaroo, — , -majitaroo.
VII. Total Variables.
(1) The i-counterpart fornr
ki, — , — .
(2) The zero-counterpart
form.
- Ji,
(3) The re-counterpart form
kure, sure, -ue.
(4) The a-i-counterpart form
— , — , -zu.
(5) The a-e-counterpart form
ko, se, — .
(6) The yoo-counterpart form
koyoo, Jiyoo, — .
(7) The ru-counterpart form
kuru, suru, -n.
(8) The te-form
kite, Jite, — .
(9) The ta-foim
kita, /ita, — .
(10) The tari-form
kitari, Jitari, — .
(11) The tara-form
kitara, Jitara, —
(12) The taroo-form .
kitaroo, Jitaroo, — .
The Second Base.
Renyokei.
The 1st Base, 2nd Base, 5th Base.
Mizenkei, Renyokei, Meireikei.
The Fourth Base.
Jokenkei.
The First Base, The Second Base.
Mizenkei, Renyokei.
The First Base.
Mizenkei, Meireikei.
The Third Base.
Sh ash ikci, Renta ikei.
Gerund.
The Past Tense Form.
The Alternative (Frequentative)
Form.
NEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1047
As can be seen, the proposed classification divides the constituents ^
of spoken Japanese into two large groups : Invariables and Variables.
^By Variables I mean those constituents which may assume distinct
forms in different verbal contexts,^ and by Invariables those which
do not.
The first large group of Invariables is Connectibles. i.e. the con-
stituents that can be connected to each other or to Variables with or
without the aid of what are here called Supplements ". This sub-
group comprises Nouns, Pronouns, Numerals, and Adverbs of the
conventional classification. The term " Numeral " is sufficiently clear
on the whole, but the remaining three — " Noun ”, " Pronoun ”. and
'■ Adverb ” — leave ample room for dispute. No one would object to the
term “ Noun " applied to such constituents as inu (dog), tori (bird),
hombako {bookca.se). te (hand), and kin (gold). But when a certain
form of Verbs, like iki (go), amari (excess), and Jirase (inform), is called
a Verbal Noun or Juniaigen (Quasi-substance-word), we are made to
understand that a constituent can be at once a Noun and a ^Vrb.
which is, of course, absurd. Similarly baka (fooUshness. fool), bikko
(lameness, cripple), etc., satisfy the definitions of both Meishi (Nominal
words) and Jotai no Fukiiski (Adjunctive words of state). This is
exactly where qualificative terms lose their precision. In morphology
we are not in the least concerned with the semantic aspect of the
constituents.^ Whether or not a given member is the name of a thing,
an action, or a (juality is quite immaterial. All that we need is to
perceive what a given constituent sounds or looks like, and what
position it always occupies in connected speech or writing. And in
conformity with its external appearance or the position it always
occupies in relation to other members a discriminatory name may be
given to it for convenience.
The name ” Connectible ”. though somewhat awkward and
unattractiv'e, is the be.st I have been able to inv^ent for this category
with an e.xtremely large number of constituents including the whole of
^ The name ’■ Con>titiient " is here jriven to a eonstantly recurring phone sequence
who'^e components are so firmly joined together that any attempt to divide it into
smaller parts would either cause unnecessary inconvenience or serve no useful purpose
in a study of the language under consideration. A constituent may t>f course consist
in a simile phone like a, e, n, etc.
- Some Invariables may occur m more than one form, e.g. mina and m inna, But
these are always found in precisely the same verbal contexts, and so outside the
category of Variables.
Cf. “ The Technique of Semantics " by ,1. R. Firth, in the Philological Society's
Transactions^ 1935, pp. 48-9.
1048
S. YOSHITAKE —
the unadulterated Sino-Japanese elements. If desired, however, the
Connectibles can be divided into five groups according as they can or
cannot be directly followed b_v ga, no, and na (plus a constituent _
other than no and n). These differentiae may best be illustrated
follows : —
qa
no
na
(1) qa-no-Connectibles .
. Yes
AYs
No
(2) qa-na-Connectibles .
. Yes
Yes
A'es
(3) no-Connectibles
. No
A"es
No
(4) na-Connectibles
. No
No
A'es
(5) Direct Connectibles
. No
No
No
In this way it is possible to separate Connectibles like baka aiui
bikko from inu, hombako, etc., on the one hand, and from dgoobu,
taihen, etc., on the other. Likcwi.se, the two kinds of the so-called
Adverbs can be kept apart as .shown by (3) and (5). Moreover, the
propo.sed classification enables us to distinuuish the Adjectives kono.
sono, ano, and done, from the corresponding Pronouns kore, sore, are,
and dore. If. following the conventional morphology we name the
former set " the Attributive form of Pronoun ", then what form ot
Pronoun are we to call the latter ? ” The Pronominal form of
Pronoun “ would certainly l)e mo.st disconcerting.
The Supplements are divided into Initials. Finals, and Movables ^
according to the positions they occupy in a phonetic .se([uence. Tlie
Initials .stand at the beginning of a phonetic sequence, and compri.se
hai, ha, hee, he, e, q, ie, iie, iya, aa, a, saa, sa. yaa, ya, oi, kora, sora,
oya, ara, maa, naani, moji, JikaJi, yareyare, o-, go-, dai-, oo-, ko-, o-.
me-, etc. The Finals are found at the end of a phonetic sequem e.
embracing yo, i, e, ro, tomo, ze, zo, zoo, sa, na, -san, -sa, -mi, -Ji, etc.,
and the .so-called Numeral Cla.ssitiers -nin, -hiki, -hai, -hon, -soo, etc.
The ^Movables may occupy different positions according to circum-
stances. and comprise wa, ba, mo, qa, no, o, ni, e, de, yori, kara, made,
sae, koso, dake, Jika, bakari, hodo, qurai, to, demo, nado, nandzo, yara,
ya, Ji, naqara, dsa, d3aa. 3a, 3aa. keredo, kedo, nee, ne, -tte, etc.
Both Initials and Finals often provide a useful means of picking out
that part of a phonetic se(juencc which is commonly known as direct
quotation. However, since the Supplements comprise only a limited
number of constituents,^ it .seems hardly necessary to divide them into
smaller groups. For practical purposes it would be sufficient to make
' This name has been suggested by Firth,
Their number is well under half that of Irregular Verbs in English.
NEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1049
a list of them and state that to these Inrariahles we give the name
“ Supplements
Now for the Variables. These are divided into Part Variables and
Total Variables under the proposed classification. By Part Variables I
mean the Variables which contain an invariable element, as contrasted
with Total Variables which do not. When the invariable part of a
Variable can be treated as a constituent it is here called “ the zero-
form ’■ of that Variable. The form ending in ku and built upon a
zero-form is here named “ the ku-form When a Variable ending in
a cannot be treated as a zero-form it is here called “ the a-form ".
W’ith these three forms as criteria it is possible to distinguish
the following three types of Part Variables.
(1) a-Variables ; Variables that have an a-form but not a zero-form.
(2) zero-Variables : Variables that have a zero-form but not a
ku-form.
(3) ku-Variables ; Variables that have a ku-form.
To the remaining nine Part Variables na, da, desu, -masu, nasaru,
kudasaru, gozaru, ira/Jaru, and o/Jaru are given the name " Shortened
Variables ".
The a-Variables include not only the so-called Consonant Verbs
but also -qaru, -burn, etc. Similarly the zero-Variables contain -reru
(-rareru) and -seru (-saseru) besides such constituents as miru (see)
and deru (emerge). It is unprofitable to distinguish Variables of the
type miru from those of the type deru, as is done in Japan, since they
both change their forms in e.xactly the same manner. Likewise the
term ” Consonant Verbs '' is not at all satisfactory, because the
category in question includes such Variables as iu (say), au (meet),
ou (drive away), which do not contain a single consonant.
As shown on pp. 1042-4, the a-Variables have ten distinct forms
and the zero-Variables nine. Every one of the forms belonging to
either of these two groups of Variables morphologically corresponds to
one or more forms belonging to the other group. To take an c.xample.
deru may find a place in the paradigm scatter ^ de, dere. deru, dete,
deta, etc., but only the zero-form de can be directly followed by -masu,
thus demasu.- Similarly, kasu may find a place in the scatter ka/i, kase,
kasa, kasu, kajite, kajita, etc., but onlv the i-form kaji can be
* For this term see Firth s review, op. eit., p. 1 Ig : ■ The Technuiue of .Semantu s. "
Up. oit., p. 62.
- Such a phonetic sequence as detemasu (is out ) i> here treated .us a variant of dete
imasu and not as a junction of dete and -masu.
1050
S. YOSHITAKE —
immediately followed by -masu, thus kajimasu. The same applies to
the zero-form of all other zero-Variables and to the i-form of all other
a-Variables. We may therefore state that the i-form of a- Variables is
a morphological counterpart of the zero-form of zero-Variables. But
the zero-form of zero-Variables may also be directly followed by zu,
which cannot immediately follow any form of a- Variables except the
a-form. e.g. dezu, kasazu. Moreover, the zero-form of zero-Variables
(with the exception of those which are derived from the a- Variables
e.g. yomeru) may be directly followed by ro, whose ecjuivalent yo
cannot directly follow any form of a-Variables except the e-form,
e.g. dero, kaseyo. On the other hand, the e-form of a-Variables may be
directly followefl by ba, which mav also directly follow the re-form,
but not the zero-form, of zero-Variables. e.g. kaseba, dereba. This
means that, while the a-form. the i-form, and the re-form are all
univalent, the e-forni is ambivalent and the zero-form trivalent.
The u-form amt the ru-form are characterized by the fact that they
are the only forms of a- Variables and zero-Variables that can be
directly followed by the prohibitive Final na. They are therefore
univalent counterparts. So are the oo-form and the yoo-form. since
neither of these can be replaced by any one of the forms mentioned
above, while they both are found in the same verbal contexts. The
following table will show the morphological ecjuivalence under
di.se u.ssion ; —
a-Variables
zero-Variables
1
/
The i-form
The zero-form
The e-form
The re-form
The u-form
The ru-form
The oo-form
The yoo-form
The term ' Base used by McGovern and others is suitable only
for distinguishing something from its derivatives. When applied to
different forms of the same Variable under the categories of First,
Second, Third Base, etc., it loses its suggestive value. Such a method
of classification should be avoided where possible. Indeed, even the
negative criteria as adopted in dividing the Connectibles are le.ss obscure
NEW CLASSIFICATION OF CONSTITUENTS OF SPOKEN JAPANESE 1051
than the numerical classification. The Japanese practice of ascribing
two or more names to one and the same form or category is likewise
undesirable.
The three constituents kuru (come), suru (do), and -n (not) contain
no invariable elements and are therefore called Total Variables. ^ It
is quite wrong to designate kuru and suru “ Irregular Verbs ”, for we
find no irregularity in either of them. The\' change their forms just
as regularly as other Variables, and so we cannot single them out as
“ Irregular Verbs ” any more than we can de.scribe an unsociable crank
as an irregular fellow.
However, the Total Variables change in appearance to such an
extent and so differently from a-Variables and zero- Variables that
some of their forms do not share any common element. It is therefore
not easy, by means of their terminations alone, to differentiate all the
varied forms, either among themselves, or from the a- Variables and the
zero-Variables. Nevertheless, there is a definite correspondence, for
example, between ki of kuru and the i-form of a- Variables, for they are
the only forms, of these Variables that can be directly followed by
-masu. And since the i-form is univalent, ki may be called the i-counter-
part form of kuru. For similar reasons kure and sure may be named
the re-counterpart form, kuru and suru the ru-counterpart form, and
koyoo and Jiyoo the yoo-counterpart form, while Ji may rightly be
designated the zero-counterpart form. Both ko and se, and no other
forms of kuru and suru, can be immediately followed by zu, and
therefore correspond to the univalent a-form. But, like the e-form, they
can also be directly followed by yo or its variant i, thus koi and seyo.
Although the e-form is ambivalent, being a counterpart of the re-form
as well as of the zero-form, since the name the re-counterpart form has
been given to kure and sure, we may without any ambiguity designate
ko and se the a-e-counterpart form. Lastly the ambivalent zu must be
called the a-i-counterpart form. Thus we arrive at the seven forms
(l)-(7) as tabulated on p. 1046.
The Shortened Variables include the i-e-couuterpart form and the
a-e-counterpart form. These names are, by elimination, given to the
two ambivalent forms as in the case of the a-e-counterpart form of
Total Variables.
The te-forms are self-explanatory, for the same termination is
utilized as a criterion for all Variables, The voiced variant -de, as
found in the a- Variables, is a modification of its voiceless prototvpe, the
* This name has been suggested by Firth.
1052
S. YOSHITAKE
change being brought about by the preceding consonants that are either
nasal or readily susceptible of nasalization^ The te-forms are not
recognized as forms by native grammarians, because the ending te
was at one time a Total Variable freely afiixable to the i-forms and the
zero-forms of a-, zero-, and Total Variables, as also to the ku-form
of ku- Variables, and still remains so in the written language. But since
in modern colloquial -te no longer survives as a separable element, it
may be made use of as a discriminating factor for that particula'
form of A’ariable which is, .strictly speaking, neither a Gerund nor .i
Participle. The same is true of the endings -ta(da), -tari(dari), etc.,
which serve as formal criteria for those forms to which no appropriate
names have hitherto been given.
In addition to tho.se tabulated on pp. 1042-6 there are forms ending
in -utte (like urutte, in contrast to utte), -rntte (like mirutte, in contrast
to mite), and -tatte or -datte (like uttatte, mitatte, yondatte). But
these may best be treated as combinations of the u-form. the ru-form.
and the ta(da)-form with the Movable -tte. Similarly the forms ending
in -tja or -t/aa (hke utt/a, mitja, uttjaa, mit/aa) and -d 3 a or -dsaa
(like yondsa, yond 3 aa) should be regarded as the variants of the
te(de)-form -t- wa. Likewise, -teru (deru) is a variant of the te(de)-
form -f iru, and -t/atte (d 3 atte) of the te(de)-form -r /imatte. To
be treated in similar manner are -t/ai (d 3 ai), -tjae (d 3 ae), -t/au (dsau),
-tjatta (d 3 atta), -tjattari (dsattari), -tjattara (dsattara). -tjattaroo
(dsattaroo), etc.
Little need be .said of the various forms of ku- Variables. It is quite
unreasonable to overcharge the form (2) with three distinct names,
whereas the forms (5) and (6). having different morphological functions,
should deserve separate names. To the eleven forms given on pp. 1044-5
may be added a kare-form like yokare (be it good), afikare (be it bad),
although not all the ku-Variables have this particular form.
To sum up, the proposed cla.ssification is not onlv clearer than the
conventional schemes, but it also contains a considerably smaller
number of nomenclatures, .some of which are bv no means new. The
term "Variable is sugge.sted by the Japanese terminology Yofjfti
(Inflective words) and Ddsfii (Variable words). The names " i-form
and ■■ ku-form , applied to the ku-Variables. were emploved by
' -nde is the j)hunetic' unplKatioii of -nte, -mte, and -bte; -ide of -gte, and -tte
of -tte, -rte, -kte, and -fte. If this latter .system of notation, which is expressive of the
morphological consonant junction.'*, t)e adopted, the te(de)-form mav .simply be called
e te-form. Cf. The Technique of .Semantics op. cit., pp. .59-60.
NEW CLASSIFICATION OP CONSTITUENTS OF SPOKEN JAPANESE 1053
McGovern more than fifteen years ago, while the appellation “ Final ”
as used here covers a greater variety of constituents than in Yamada’s
classification.
In selecting my terminology I have for obvious reasons endeavoured
to confine myself to everyday language, and nomenclatures similar
to the ones adopted here will easily be found in colloquial Japanese.
As a specimen I suggest the following. Alternatives are given within
parentheses.
‘ Invariables . . Henlei shinai mono.
Connectibles . . Tsunagi-eru mono.
qa-no-Connectibles . qa mata ira no wo IsHzuhete tsunagi-
eru mono.
qa-na-Connectibles . ga mata wa na wo tsnzuh’te tsunagi-
eru mono.
no-Connectibles. . no tco tsuzidete tsunagi-eru mono.
(qa mata na na wo tsuzukete wa
Isunagi-enai mono).
na-Connectibles . na wo tsuzukete tsunagi-eru mono.
(qa mata wa no wo t.suzuke(e wa
tsunagi-enai mono).
Direct Connectibles . Chokusetsu ni tsunagarcru nmio.
(qa, no, mata wa na wo t.suzukete wa
tsunagi-enai mono).
Supplements . . Tsukekuwaeru mono.
Initials . . Hajime ni tsukekuwaeru mono.
Finals . . Owari ni tsukekuuaeru mono.
Movable.s . . . Iroiro notokoro ni tsukekuirueru mono.
Variables . . Henkei suru mono.
Part Variables . . Ichibubun (lake henkei suru mono.
a-Variables . a.-Kci no aru mono.
zero-Variables . . zero-Kei no aru mono.
ku-Variables ku-AVi no aru mono.
Shortened Variables . Mijikaku natte iru mono.
Total Variables . Sukkari henkei suru mono.
Each of the regional dialects, the professional and class dialects,
and the vernaculars of women and children, as well as the VTitten
languages of different times must be examined severallv, and its con-
stituents classified on similar hnes as here proposed. Until this has been
done it is almost impossible to venture anidhing like a complete
^ historical morpholog\' of the Japanese language.
The Structure of the Chinese Monosyllable in a
Hunanese Dialect (Changsha)
By J. R. Firth and B. B. Rogers
^TIHE only kind of speech behaviour with which tlie present study
is directly connected is the oral naming of Chinese characters.
Chinese characters have been prominent cultural objects for thousands
of years. Even the sketchiest description of what they have been
called by countless milhons over a vast area of Asia would be a colossal
task. All that is here attempted is a systematic analysis of what a
certain number of selected characters were called by Mr. K. H. Hu,
of Changsha.
If a precedent must be given, let it be Adam in the Garden. Created
things were brought before him to see what he would call them.
They did not come into his world until he had called their names. And
the names by which he called them were a new creation, henceforth
part and parcel of his world. It might even be said that they did not
exist until he had called their names. A great deal of grammatical
phonetics is concerned with the de.scription of the spoken names of
written words and tells the uninitiated what to call them when they
are faced with them or when they want them. Such information can,
of course, be used indirectly in continuous speech. But we would
emphasize once more that the analysis here presented is not directly
concerned with what is properly called general speech behaviour.
Consequently no connected text is given. The notation, however,
could be used as a simple Roman orthography, and lends itself to all
modern printing de\dces. The ordinary tj’pewriter keyboard could be
used, and for telegrams, in this dialect at any rate, the ordinary tele-
writer could be used and the sending of such messages very much
simplified as compared with the present numerical code method. A
sample telegram is given at the end.
Lastly, the use of the word monosyllable in the title and in the
text does not imply that Chinese is to be classed as a monosyllabic
language. In actual speech reduplicative and dissyllabic elements are
quite common.
1056
J. R. FIRTH AND B. B. ROGERS
The description of the pronunciation to be associated with the
notation employed is also sufficient for practical purposes. The principle
followed is that the main diacritica of the various types of syllable
should be described and simply recorded. These diacritica may be
regarded as occurring in two places, first place or initial position, and
second place or final position. The tonal diacritica need not be
“ placed ”, although the notation employs letters in final position.
These are not, therefore, counted as being in a “ place
Tones
The technique indicated above was also applied to tones. The
analysis and orthographic representation of these is based on
observations of the tones used by Mr. Hu when called upon to name the
selected characters placed before him. The tonal beha\'iour of such
syllables in connected speech was not investigated, except for the low
falling variant of tone 4 often used in certain syllables in connected
speech. The nature and function of the so-called “ tones ” of Chinese
cannot be understood or econoraicaUy represented in orthography,
until a thorough study of types of sentences in general speech behaviour
has been completed by enlisting a number of workers.
Figure 1
Relative Pitch and Length of Tones
1
2
3
4
5
■
Mid- falling
Mid high-
rising
Low mid‘
rising
Mark
y and w
h
V
unmarked
doubling
and 0
lab
sob
fab
—
me
______
X
t
ray
'■ X
r
7
dob
/I
le
y
—
. /I
lab
sob
m
JiML
itiah
mav
ms
maa
mother
bemp
horse
1
scold
wipe
* See Fu Liu, Lea mouvementa de la langue nationale en Chine, 1925, paras. 125-7,
p. 35, and also paras. 171-7, pp. 49-50.
STRUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1057
Relative Frequency of Tones
Three types of count for frequency were made, the first from a
normal personal letter of 802 words (A) ; the second from a classified
list of the 1,013 possible syllables in Changsha dialect (B) ; and the
third from a national “thousand-character” list (C). The resulting
figures are shown in the following table and graph ; —
Tone
(A)
(B)
(C)
1st
. 143
227
229
2nd
. 128
168
219
3rd
. 199
248
200
4th
. 225
257
334
5th
. 107
113
189
802
1,013
1,171
No. of
words
Figure 2
1058
J. E. FIETH AND B. B. EOGEES—
From the graph it will be seen that there are reasons for thinking
that the fourth tone occurs most frequently. Therefore this tone is
left unmarked, the final letters, y, w, h, v, o, and the doubling of the
final vowel letter, being used to mark the other tones.
In the marking of the first tone y is used with i, e, ei ae, en, eun,
and w with a, o, eu, u, ao, ou, on, aon, an, un. In the marking of the
fifth tone the doubling of the final vowel letter is satisfactory for all
Table I
Tones — "With Vowels and Correlative ArrRin'jrE.s
Tone 1, marked
with y and w.
iy, ey, aw, ow, euw,
uw, eiy, aey, aow,
ouw, eny, euny,’
onw, aonw, iny,
anw, unw.
Long — longer than 3. Nor-
mal voice quality. Some-
times ends with slight
creak.
Tone 2, marked
with h.
ih, eh, ah, oh, euh,
uh, eih, aeh aoh,
ouh, enh, eunh,*
aonh, onh, inh,
anh, onh.
Long — often longer than 1.
Voice quality breathy,
hollow, “ chesty " with
.slight initial friction.
Tone 3, marked
with V.
iv, ev, av, ov, euv,
uv, eiv, aev, aov, ,
OUT, env, eunv,*
onv, aonv, inv,
anv, unv.
Medium length — shorter
than 1 and 2. Clear
“ head voice — vowels
usually closer and clearer.
Slight final creak. Clo.s-
ing na.salization slight.
Tone 4, un-
marked.
i, e, a, 0, eu, u, ei, ae,
ao, OU, en, eun,* on,
aon, in, an, on.
Short. Clear " head " voice.
Ends with check. Closing
nasalization very slight.
Final n very short and
checked.
Tone 5, marked
with doubled
i, e, a, 0, and
final 0.
ii, ee, aa, oo, euo,
uo, ouo.
IMedium length — with the
a-vowel, longer than 3.
Slight initial breathiness.
Checked. No closing
nasalization.
^ See note>> on Table VI, p. 106o.
STRUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1059
letters except u. To avoid ambiguity with n in writing, final o is used
instead of doubling the u.
In conjunctive spelling which would link syllables together to form
words it would sometimes be necessary to use a hi'phen after syllables
in tones one and two to distinguish the final w, y, and h, which are
tone letters, from the initial w, y, h, which represent differences of
pronunciation. The hyphen might also be necessary occasionally after
tone five.
Notation
Vowels
The number of letters required for the symbolization of vowels is
five, as follows ; i, e, a, o, u. This is not to say that there are only five
vowels, as will be clear from the tables.
Consonants
The number of letters required for the symbolizing of consonants
is seventeen, as follows (in alphabetical order) ; b, c, d, f, g, h, j, k,
1, m, n, p, s, t, w, y, z.
Tones
Only one letter in addition to the above is found to be necessary
to mark tones ". namely v in final position. The other devices for
representing the prosodic diacritica include the use of y, w, h, and o
in final position, and the doubling of the final vowel letter. In contra-
distinction from the above positive marks, the fourth tone, which
occurs most frequently, has zero mark.
Phonetic Analysis
As speech behaviour, the naming of the characters is just one
complete act, a configuration of bodily postures and mm ements not
easilv dissected. But the difterences between these oral names can
be systematicalh' described, classified, and represented in notation.
To do this it will be found convenient to regard the Hunanese mono-
svllable as having one, two, or perhaps three places in which the
phonetic diacritica may bo said to occur. In these places various
nlternances have differential function. (See alternance tables.) The
tonal diacritica and possibly also what we have called yotization and
libio-velarization mav be considered as syllabic features.
VOL. Vin. P.4RT 4.
6S
1060
J. E. FIRTH AXD B. B. ROGERS —
The follo^’ing vowels constitute a single term alternance in one-
place syllables — (tones apart) : —
i, 0 , eu, u.
Two-place syllables are by far the most numerous, the initial
alternance being consonantal and the final alternance vocahc, including
closing nasahzation. In syllables with final nasal, onlv three
differentiations of vowel quality occur — fairly close front, mid-neutral,
and open central. That is to say the differences between i, e, ei, which
operate in other contexts are here neutralized. Similarly with o, eu,
and u. hen therefore we symbolize these three syllabic element.s
thus — in, an, un, we do not imply that the three vowels are variants
of vowels Xo. 1. Xo. 3, and Xo. 6 respectivelv. ^\'e tlo not so regard
them.i
In addition to the classification of Hunanese monosyllables
according to the number of “ places.”, it is convenient to distinguish
them also according to certain phonetic characteristics as follows : —
Syllables (i) with final nasal
(ii) with closing nasalization
(iii) with yotization
(iv) with labio-velarization
(v) with yotization and nasalization
(vi) with labio-velarization and nasahzation
and negatively
(vii) without the above diacritica.
Syllables with final nasal only show three differences of vowel
quality, i-like, a-like, and a-like, whereas closing nasalization is
associated with four qualities, the nasalization in en and eun being of
a front quality, and in on and aon of a back quality.
The distinction between yotized and velarized syllables is a striking
contrast of resonance following the initial consonants, one front,
rather like i with .slight spreading of the lips, and the other back rather
like an unrounded o or o with neutral lips or slight inner rounding.
There must be no pouting. This contrast is a broad distinction of
front and back resonance made use of in a similar way by manv diverse
languages. In the dialect we are considering, the yotization and labio-
velarization differences may be regarded as syllabic diacritica and not
as being " placed
^ Table VII and notes.
^ See pp. 1059, 1073.
STEUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1061
Vowel Alternances
Simple
The simple vowel qualities may be suggested by the following
symbols and diagram : —
Table II
Number of vowel .12 3 4 5 6
Phonetic notation .1 e -a, a, o, t ui u, z,
a 3 , etc.
Orthographic notation i e a o eu u
Figure 3
Xoles
General : A'owel quality, diphthongization, voice quality, length,
and final check ” or creak vary with the tones ; that is to say, they
are correlative attributes. See Table I.
A’owel No. 1 — i — close and not diphthongized.
Vowel No. 2 — e — varies in quality ; sometimes starts with an
i-like glide, and sometimes shows slight closing diphthongization,
chiefly with the 3rd tone.
'I’owel No. 3 — a — three variants are shown in Table III.
1062
J. E. FIRTH AND B. B. ROGERS —
Vowel No. 4 — 0 — generally of constant quality, except with
the second, third, and fifth tones when there is slight closing
diphthongization. In yotized syllables in which this vowel occurs the
y is more than usually i-like.
Vowel No. 5 — eu — a more centralized variant is used in labialized
syllables.
Vowel No. 6— u — varies considerably according to context. See
Table IV.
Table III
a
Back ^
ilid 2
Forward ®
te
pa
T
hya
yaw
fear
under
raven
A
baa
cyaa
yah
eight
fitting
tooth
taw
jyaw
yav
he
family
elegant
iz
da
ya
large
second
mav
yaa
horse
repress
iSiJ
shwaa
brush
Notes
(1) The commonest quahty of a is back, fairly near cardinal [a],
after the consonants given in the column below, and also after the
following consonants : k, f, w, 1, s, ts, dz, kw, gw, and jw.
STRUCTUBE of CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1063
(2) After (the consonants) hy, cy, and jy the position of the vowel
a is slightly advanced.
(3) After initial y the position of a is further advanced to one
approximating English [a]. It will be noted that the five examples given
in this column vary in tone only.
Table IV
The Syllabic Element U
m and u 1
5 .'
With slight Fronted Alveolar syl- Post alveolar Retroflex
bilabial friction unrounded tabic velarized -syllabic velarized syllabic velarized
phm
puw
A
jtt
yuo
fB
sz
su
cuo
shn
spread
enter
four
foot
TE
be
bm
bno
it:
njii
nyuv
it
tsz
tsuv
m
•lis
ju
a
zuo
not
woman
this
arrive
P
sun
kuw
shynw
dznh
khm
'
gjii
m
dzz
dry
book
style
gnv
C3ru
gin
*
tfii
ancient
go
fu jyuh
^
rich
exclude
wuh
^ j win
I without
Vowel No. 6 — n — : Notes
( 1 ) This common syllabic element is usually produced with friction.
After p, b, k, g, f, and w the “ vocalic ” component is back, half
close, unrounded, accompanied by bilabial friction, especially at the
sides of the mouth. After p and b there is sometimes a short bilabial
trill.
1064
J. E. FIRTH AXD B. B. ROGERS
(2) After y, and in votized syllables a close centralized vowel
with slight friction.
(3) After s, ts, dz, c, j, sh, and z, the syllabic element is the voiced
homorganic continuant, velarized. with reduction of friction.
Diphthongs
The diphthongs mav be suggested bv the following symbols and
diagram : —
Table V
Number of vowel . . 7 8 9 10
Phonetic notation . . ae ae I qy at
Figl're 4
Xofl'S
Vowel No. 7 — ei — narrow centralized diphthong.
\ owel No. 8 — ae — after sh and c often begins with an T-like
glide.
4 owel No. 9 — ao — . No lip-rounding ; this vowel occurs in
all types of syllable except those wdth initial w, (hw) f, and [?] hy,
STRUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1065
though not with the fifth tone. See note on maximum consonant
alternance, p. 1073.
Yowel No. 10 — ou — . No lip-rounding. Back unrounded glide
from advanced half-open position to slightly less advanced half-
close position. In yotized syllables the y element is i-like, and the
^'ariant of ou used is rather like y.
Table VI
Closing N.asaliz.ation
o 1
4’
- 1
e
0
Y
m.
QT
sey
e
sow
0
seuo
eu
saow
ao
11
--
12
13
14
ee
00
m
91
DY
seny
en
sonw
on
sheun
eun
saon
aon
Notes
The simple vowels e, o, eu, and the diphthong ao have nasalized
correlates. The nasalization difference in the case of en, on. eun. affects
the end of the vowel, and gives the impression of a closing diphthong
with a nasalized ending. This differentiation may therefore be termed
chs'nuj nasalization, firstly because in en and on the lowering of the
^■elum appears to be associated with a clo.ser vowel quality and secondly
the nasalization is only associated with the end-phase of the vowel. In
the case of eun the vowel quality is not only fronted at the beginning,
but moves in the direction of i in the closing nasalization. The end
phase of the diphthong ao is not rounded, but i.s back, somewhat
centralized. This diphthong has its nasalized correlate aon. The latter
moves within a similar tamber range though it begins and ends some-
what closer than the unnasalized correlate ao. In yotized syllaldes the
i)eginning of the diphthong in aon may be centralized.
It would be possible in orthography to dispense with eun as it
represents a specific closing nasalization which occurs only after sh,
j, and c. In these syllables the difference between eun and en is
immaterial or neutralized, so the notation en would be unambiguous.
1 See Table II.
- See Table V.
1066
J. E. FIRTH AND B. B. ROGERS —
Syllables with Fixal n
Only three vowel differences occur in syllables with final n ; —
Table VII
Number of vowel
1.6
16
17
Phonetic notation
in
an
an
sen
Orthographic notation .
in
an
un
Xoles
Syllables with final clear alveolar n, with no vocalic off-glide. In
syllables with a final n there are only three taniber differences : (1) an
i-like vowel lowered and retracted ; (2) an a-like vowel, more front
than vowel No. 3 and in yotized syllables rather se-like and centralized ;
and (3) a neutral vowel half-open. Our use of i, a, and u in the
orthography in, an un, is not to be taken as identifying these elements
with Vowels Nos. 1, 3, and 6 in other contexts.
Initial Consonant Alternance
It is convenient to classify the initial alternances first of all
according to the number of essential articulation differences for each
of the three types of consonant ; (a) plosives and nasals ; (6) fricatives,
and (c) affricates.
Secondly consonant terms are then multiplied by the following
differences : (a) the aspiration-tensity difference, and (6) the voice
difference.
Thirdly the syllable pattern is further differentiated by diacritica
which are here termed yotizatioii and labiovelarization. Hitherto
most scholars have regarded these differentiations as part of the vowel
system, but analogous phenomena in Burmese suggested it might
make for clearer analysis to treat this differentiation of the mono-
syllable by grouping the contrasted “ y "-like and “ w "-like elements
with the consonantal terms of the initial alternance. It should be noted,
however, that in some contexts the y-element is more vowel-like in
quality, in others more consonantal.^
* See pp. 1062, 1070.
STEUCTUBE OF CHINESE MONOSYLLABLE IN A Ht’NANESE DIALECT 1067
Plosives and Nasals
For plosives and nasals there are three essential articulation
differences (not including variations consequent or dependent on the
yotization and labialization differences), bilabial, dental, and velar.
As a practical convenience in the table a separate column shows the
pre- velar articulation of n and ny and the palatal articulation of
gy and ky, which in the plosives correlates with another difference, the
yotization difference.
As basic terms for this alternance we take b, d, g. These three
consonants are rather like whispered b, d, g — that is to say, they are
not really voiced although there is obviously some associated laryngeal
and infra-glottal behaviour contrasting with the different chest and
larynx behaviour associated with the aspirated correlates.
These three basic articulations are differentiated by four further
differences, which we now associate with the initial consonantal
alternance, in continuation of the above classification of syllables.
i. The aspiration difference.
ii. The yotization difference.
iii. The velarization difference.
iv. The nasalization difference.
i. The Aspiration Difference
The three voiceless stops written p, t, k, are released with fairly
strong aspiration, more than would be heard in Southern English in
the case of initial p, t, k, followed by a vowel in a stressed syllable,
but not so strong as in Indian languages. So far then, we have six
stops.
ii. The Yotization Difference
To tlie above six stops, sLx more are added by the yotization
difference, viz. by, py, dy, ty, gy, ky.
Notes on ky, gy, c, j, cy, and jy
(1) Before ao, ou, and an the- difference between ky, and c, gy
and j, is significant ; also before o and aon in the case of gy and j.
(2) Before i, in, e, en, the pronunciation of k is palatal, and the
difference between k, ky, and c is immaterial. In many syllables either
the aspirated palatal plosive or the aspirated affricate may be used.
The use of gy and j as alternative pronunciations in similar contexts
is not quite so common, but it does occur.
1068
J. R. FIRTH AXD B. B. ROGERS
(3) In syllables containing the syllabic element u, ky and gy
do not occur, but the difference between c and cy is significant. From
this fact and other variant pronunciations of ky it would seem that the
difference between ky and cy, gy and jy is likely to be immaterial.
Table VIII
The Differextiatiom op Velar axd Palatal Plosives axd
THE Palato-alveolar Apfricates
k
C
ky
cy
g
j
gy
jy
i
kiy
— ,
(kyiy)
giy
(gyiy)
—
m
e
(kee)
—
kyee
—
(gee)
gyee
a
kav
kyaa
--
—
—
gyaw
0
kow
—
kyoo
gow
joo
gyoo
---
#4
in
¥
m
eu
keuo
ceuo
—
--
geuo
jeuo
--
$
m
m
u
kuw
cuw
-
cyuw
guw
juw
■ -
jyuw
m
ae
kaey
—
—
gaey
—
—
—
m
ao
kaow
caow
kyaov
gaow
jaow
gyaov
tV
t-ij
is
Mi
ou
kouv
couv
kyouo
gouv
jouw
gyouw
-
□
m
m
ni
en
—
—
kyeny
—
—
—
gyeny
M
m
on
konw
—
—
gonw
-
%
eun
—
ceunv '
-
—
jeunv ’
m
aon
kaonw
eaonw
gaonw
jaonw
gyaonh
- -
$
%
X
4*
m
in
—
(ciny)
kyiny
-
(jiny)
gyiny
m
rfi
an
kanw
canw
kyanw
ganw
janw
gyanw
TIJ
B
m
nn
kunw
cunw
—
gunw
junw
•
- .
ifi
m
m
m
* St-e notes on Table VI, p. 106.5.
STRUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1069
The difference between ky and cy, gy and jy is immaterial and may
be regarded as alternative pronunciations. As will be seen from
Table IV. a feature of certain syllables is the use of a syllabic con-
tinuant often homorganic with the initial consonant, when that is
fricative or affricative. This element is here symbolized by u. Before
this element it is necessary to distinguish between c and cy, j and jy,
but of course ky and gy would be unambiguous provided that an
affricative pronunciation was understood.
iii. The Labio-Velarization Difference
To the above twelve stops two more are added by the labio-velariza-
tion difference, kw and gw, making fourteen stops in all, alternating
in initial position.
In these group-plosives (py, kw, etc.) the combination of aspiration
and yotization, and of aspiration and velarization, produce
characteristic quahties in the release of the stops. These contextual
variations are noted below : —
Sates on the Aspirated Plosiees.
(1) Palatalized and g-like aspiration when followed by i.
(2) x-like aspiration when followed by a.
(3) Back resonance of aspiration wheii followed by o, and more so
in the case of kw.
(I) The aspiration of py, ty, ky, is g-like followed by i-like yotization
before ao and also sometimes before e and en.
iv. The Sasal Difference
It will be seen from the table that the nasal difference, with
yotization. adds si.x more terms to the initial alternance on the basis
of the three articulations noted at the outset.
1070
J. R. FIRTH AXD B. B. ROGERS —
Table IX
CON'SOXAXTS. PlOSIV'ES AND XaSALS
(Initial Alternance only)
Bilabial
Dental
Palatal and
Pre -velar
Velar
Orth. Phon.
Orth. Phon.
brth. Phon.
Orth. Phon.
Voiceless lax . b b
— —
g 9
Aspiration
difference p ph
t th
— —
k kh
Yotization by ' bj
difference py phj
dy dj
ty thj
gy jj
ky cj
— —
Labio-velariza- — —
tion difference
— —
— —
gw gw
kw khw
Nasal difference m m
and yotized my mj
nasals
1 naso-
ly lateral
a q
ny qj
[n] [Dl
Xotes
(1) 1 — The tongue tip articulation is dental, the sides of the tongue
are not completely closed against the teeth and the soft palate not
completely raised, so that the acoustic effect is slightly nasal. It may
be described as a naso-lateral.
ly — Similar observations apply, the articulation being palatalized.
(2) n — In the initial alternance n is pronounced as a pre-velar
nasal [g]. In the group ny it is nearer the palatal position, but never
sounds like fji]. It must be noted that the letter n in final position is
used to indicate closing nasalization in en, on, eun, aon, and a clear
dental nasal in the syllabic elements in, an, nn. No functional
identification of these various nasals is suggested.
Fricatives and Affricates
The plosive and nasal alternance, as we have seen, is based on three
articulations, that is. if we classify the pre-velars with the velars.
For fricatives and affricates, however, there are six articulations, as
set forth in Table X.
STRUCTURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1071
Table X
Consonants : Fricatives and Affricates
(Initial Alternancc only)
Labio-
dental
Alveolar
Retroflex
and Post-
alveolar Palatal
Velar or
Post-AYlar
Voiced
—
—
z U] —
—
Breathed
f
s
sh [g] (hy)[QorA]
h[x]
Yotized .
—
sy
shy hy [g or ji]
—
Labialized
(hw)
shw —
—
Voiced lax
—
dz
j —
—
Aspirated tense
—
ts
c —
—
Yotized .
—
dzy
jy —
—
tsy
cy
Labialized
—
—
jw —
CW
Notes
(1) The alveolar sibilant is differentiated by yotization only, giving
two terms ; s and sy.
(2) The alveolar affricates are differentiated by the voice-aspiration
difference and the yotization difference, giving in all four terms :
dz, ts, dzy, and tsy.
(3) The retroflex articulation is the base for four terms: (1)
voiced z, (2) breathed sh, (3) breathed velarized shw, and (4) breathed
yotized shy occurring only before the syllabic element u.
Initial z is much more like a retroflex voiced sibilant than the
corresponding sound in the dialect of Peiping which is sometimes
described as a sort of post -alveolar r-sound.
(4) The palato-alveolar affricate articulation is the base for a
complete series of six terms differentiated by : — ■
i. Voice-aspiration difference j c
ii. Yotization jy cy
iii. Labio-velarization jw cw
1072
J. B. FIRTH AXD B. B. ROGERS —
Table XI
Semi-Vowels axd
Fricative
Correlates
X 1
w
back semi-
vowel without
lip-rounding
f (hw)
labio-dental
or bilabial
breathed
fricative
y
semi-vowel
hy [q or ;s]
breathed
palatal
fricative
yw [Jy]
semi-vowel
group
i
—
—
yi
hyi
—
e
—
—
ye
hyee
—
a
waw
fa
ya
hya
—
0
—
—
yoo
hyoo
—
eu
—
—
—
—
yweuo
u
wu [u]
fu [fu]
yu
—
—
ei
wei
fei
—
—
ywei
ae
wae
fae
—
—
—
ao
—
—
yao
hyao
—
ou
—
—
you
hyou
—
en
—
—
yen
hyen
ywen
aon
—
—
yaon
hyaonw
—
in
—
—
yin
hyin
—
an
wan
fan
yan
hyan
—
un
wun
fun
~
ywun
Notes
The digraph hy is used for the breathed fricative correlate of y-
hw might also be used for the similar correlate of w, and indeed, it
may be so pronounced. But usually it is pronounced by making a
light contact of the inner part of the lower lip with the front of the
upper teeth, and sounds rather like f. For this reason and for other
STRUC TURE OF CHINESE MONOSYLLABLE IN A HUNANESE DIALECT 1073
alphabetic reasons the letter f is used. A bilabial pronunciation is
also possible. IVe may bear in mind that there is no surviving pw or bw.
N.B. — In the above table most of the examples given are
pronounced with the fourth tone, which has zero mark. This does not
mean that similar syllables do not occur in other tones.
Conclusion
Finally, the following single-place consonant and vowel alternances
in otherwise identical contexts should be taken as some justification of
the notation employed and of the suggestion that it might be used as
the basis of a practical Roman orthography.
(1) The maximum consonantal alternance in initial position consists
of thirty-two terms preceding the syllabic element an. as follows : —
b, p, d, t, g, gy, gw, k, ky, kw, m, 1, ly, n, ny, s, sy, sh, shw, dz,
dzy, ts, tsy, j, jw, c, cw, y, by, w, f(hw), h.
(2) The proxime consonant alternance consists of thirty-one
terms preceding ao as follows : — b, by, p, py, d, dy, t, ty. g, k, ky,
m, my, 1, ly, n, ny, s, sy, z, sh, dz, dzy, ts, tsy, j, jy, c, y, hy, h.
This proxime alternance adds by, py, dy, ty, my, z, to the thirty-two
terms given in (1), making a total of thirty-eight before an and ao.
(3) It will be noticed that in the two longest alternances given,
jy and cy do not occur. It is probable that the difference between
ky and cy, gy and jy, is immaterial. Nevertheless, j, jy, c, cy, all occur
before u. (See Tables IV and VIII.) So to our list of thirty-eight,
jy. cy, and shy must be added, making a total of forty-one con-
sonantal terms.
(4) The minimum consonantal alternance consists of the three
terms sh, j, and c, before eun. The next shortest alternance is of twelv^e
terms before on.
This latter would become the minimum alternance if eun is for
purposes of orthography amalgamated with en. (See note on Table VI.)
(5) The maximum vowel alternance is in second place after p, s,
dz, ts, and consists of sixteen terms as follows ; i, e, a, o, eu, u, ei, ae,
ao, ou, en, on, aon, in, an, un, the seventeenth, eun, occurring after
sh, j, and c. The proxime is fifteen after d, m, 1. The minimum consists
of the unique term ao after by, py, my, ty, there being a two-term
alternance of ao and ou after dy.
The diacritica of the Changsha monosyllable may thus be regarded
as occurring in two places, in the first place or in initial position, and
in the second place or in final position. The prosodic diacritica, and to
1074 structure of chixese monosyllable in a hunanese dialect
a certain extent what we have termed yotization and labio-
velarization, are characteristic of the syllable as a whole, though they
are indicated by differentiations of the final letter, by additional final
letters, by zero mark, and by the coupling of y and w with initial
consonants.
ilESS-LGE TO BE TELEGRAPHED
m ^ M ^ n
zp ®
Telegram in Ro.m.\n Orthography
IV DEUO HYOO SU HYOO WEI HYA JOUW WANV FAA FEI
KONV GI BAW IIY NYANV JOUW HOU FEIH LUNH.
PINK YOUV.
A Grammar of the Language of Florida, British
Solomon Islands
By W. G. IvENS
Abbreviations
BSOS. = Bulletin, of the School of Oriental Studies, London.
MIL. = Melanesian Island Languages, S. H. Ray, M.A., Cambridge Press, 1026.
ML. = Melanesian Languages, R. H. Codrington, D.D., Clarendon Press, 1885.
TSE. — Report of Cambridge Ethnological Expedition to Torres Straits, vol. iii,
“ Linguistics,” S. H. Ray.
excl. = exclusive, i.e. excluding the person addressed.
incl, = inclusive, i.e. including the person addressed.
pers. = person, persons.
sing. = singular.
pi. = plural.
Langdages Quoted
Bugotu = Santa Isabel, British Solomon Islands.
Lau = Mala Island, British Solomon Islands.
Mota = Banks’ Islands, Melanesia.
Sa"a = Mala Island, British Solomon Islands.
Ulawa = British Solomon Islands.
Vaturanga = Guadalcanal, British Solomon Islands.
Ti^LOEIDA is a convenient name for the largest island in the Nggela
group, British Solomon Islands. This group consists of three
inhabited islands. The smaller western island bears the native name
of Olevuga, and is separated from the other two islands in the group
by a fairly wide strait known as the “ Sandfly Passage The centre
and eastern islands are separated from one another by a narrow strait
called the “ Ututha ” by the natives, but known to white people to-day
as the “ Boli Pass ", though its older name was the " Scudamore
Passage These two islands bear the one native name of Xtfgda.
The grammar which is here presented is a grammar of the language
spoken on the two larger islands called Nggela, though the language of
Olevuga does not differ materially from the language of the other two
parts of Florida.
Bishop Patteson of the ilelanesian ilission was the first person to
make any study of the Florida language though the Spanish
discoverers recorded an odd word or two. The Rev. C. H. Brooke, of
the Melanesian Mission, was the first to make any translations into the
Florida language (the Book of Common Prayer, with certain Psalms).
* phrasc-bonk of ten page^^.
VOL. VIII. PART 4.
69
1076
W. G. IVEXS—
He worked with native assistants, and through the medium of trans-
lations in the Mota language.
Dr. Codrington published a grammar of the Florida language in
his Melanesian Languages, pp. 522-539. His information was gathered
from Florida -speaking natives in the school of the Melanesian Mission
at Norfolk Island, and also from the translations into Florida of two
Gospels and the Book of Common Prayer. He had also some native
tales written in the language. He himself helped the Rev. A. Penny
in the translation of the two Gospels mentioned above.
A fuller studv of the language of Florida has now been made
possible by the increase of translations into the language.
It will be seen from the text of this grammar that Codrington s
presentation of the verbal particles in Florida has been considerably
altered ; that the demonstratives have been added to greatly, while
the conjunctions have received more detailed treatment ; as also have
the articles, the noun.s, the gerundives, the pronouns, the verbal
particles, the verbal prefixes and suffixes, the verbs baa and vaa,
the passive, and the adverbs.
The grammar here presented has been compiled from material
found in the translation of the whole New Testament, and also of the
Book of Common Prayer, with sixty-five Psalms, in the Florida
language.
I. Alphabet
1. (a) Vowels ; a, e, i, o, ti.
(6) Consonants ; b, d, g, h, li, ngg, I, m, n, ng, p, r, s, t, v.
The vowels are sounded separately, except in cases where the use of
a doubled vowel denotes a long sound, as, e.g., when the suffixed pronoun
of the object, a, is added to verbs ending in a : gilala to know ”,
gilalaa “ to know him ” ; or in reduplications such as tiitili, where the
consonant I has been dropped. It is cjuestionable whether diphthongs
occur, though Codrington {M.L., p. 523) says that au, ao, ai. ae may
be called diphthongs in the Florida language.
The vowels have the sounds comnionlv associated with them in the
languages of the neighbourhood. The b and d are always nasalized,
i.e. they are always sounded as mb and nd. In the translations, and in
Codrington’s grammar, as well as in this grammar, there is no printing
of m before 6, or of n before d.
The sound of g in Florida approximates to the sound of g in Spanish
before e and i, or of g before a and o in Modern Greek, or of g before i
A GRAilMAR OF THE LANGUAGE OF FLORIDA
1077
in Dutch. In certain districts h is .sounded as dh. The ngg has the sound
of ng in the English word finger " ; it is a cliange from k and not
from g : nggari child ” is in Sa'a kale ; nggilii “ grave " is Sa‘a kilu ;
Xggela, the local native name of the two larger islands called “ Florida ”,
is Kela in Sa’a. In the translations and in Codrington's grammars of
Florida, Bugotu, and Vaturanga this agg sound is printed as g (italic).
Perhaps in deciding to represent the ngg sound by g (italic) in
books for native use in the Melanesian Mission it was felt that promi-
nence should be given to the double;/ sound in ngg. and that k (italic)
would convey to the eye little idea of the true sound. The use of g
(italic) = ngg is open, however, to a charge of arbitrary representation.
The ng is a nasalized form of n, and is printed n (italic) in the
translations and in Codrington's grammar and also in this grammar ;
it has the sound of ng in the English word “ singer ”.
There is a loss of m in the Florida pronouns ;/«i, ai" we " (e.vclusive).
gaii, au “ you " ; (ganii “ we " and gamu “ you " also occur). Several
Melanesian languages of New Guinea show a similar loss of in in the
exclusive form of the pronoun, Lst pers. plur. : ^lukawa kai, Wedau
at, Taupota tan-ai {TSE., pp. 430-31). There is a loss of in in the
Bugotu word koi, a noun of assemblage ; for this word occurs also in
the form koinid The pronominal forms, ai. an, cjuoted above show a loss
of g.
All syllables are open. The stress is on the penultimate syllable.
Euphony is responsible for .several peculiarities in the language ;
(1) for the omission of the article na after niiia, gana, § 3 ; (2) for the
dropping of i, u from the compound forms tai, tan, when used with
kai, kau [ta kai. ta kau) ; for the forms la ra. ka ra. tn ro, ko ro, in jilace
of tara ra, etc. ; (3) and pos.siblv for the omission of the pi. pronoun
ra when the plural has already been denoted by the suffi.xed pronoun
ra of the object, or by didira {didin) or iiggaira, § 5. That the pronoun
■na, of the object, 3rd pers. sing., is not suffixed to the preposition
ta may possibly be due to a desire to avoid the sound tana na, tatana
na being used in.stead.
II. Articles
2. (a) Demonstratives : Singular na. Plural ra na, lei, ara, ra.
(b) Personal ; a.
3. The article na is used before all nouns (except in the cases
mentioned below). It is written separately from the noun, and there is
no change of its vowel.
^ This komi is probably the Hawaiian knma “a company”, ML., pp. 231-2.
1078
W. G. IVENS —
Exceptions. — The article na is not used (1) when the genitives ni,
i, precede ; (2) with vocatives ; (3) in certain phrases, when the
connotation is general, e.g. kisu mate " to make war '' ; (1) after
nina, gana “ his ”, ruana “ second ”, na balu ” certain ”, some ' ;
siki “ one ”, “ a”, “ any.”
Na means either “ the ” or “ a ” ; the demonstrative keri may be
added to the noun when “ the ” is indicated, or to show emphasis ;
na tinoni “ a man ”, na vale “ a, any, house
The article na is used also with the gerundival forms (§ 12) ; na
niu huguanggu “ the denial of, to deny, me ” ; na bosa vaniana
“ to speak to him ” ; na lavi kehaadira “ the taking away of them ;
na dika sanianggu na raka “ my being shipwrecked ” ; ke tangomana
na vadavoraana na matana na rorodo “ can he open the eyes of a bhnd
man ? ”
The article na is used with the pronoun hava " what ? ’
“ anything ” : na hava. It is used also with the possessive nouns ni,ga :
na nina “ his ” ; na gana vanga “ his food ”.
The negative mna comes between the noun and its article na ;
see § 37.
There is an article ki “ one ”, “ any,” “ some ” ; ki tinoni any
man ” ; ki sakai “ a certain person ” ; ki totobo “ one, any,
thing ” ; ki balu “ certain people ”, ” some ” ; ki balu ganagana
“ certain ideas ”. The numeral siki “ one ” is used as meaning “ a” '■
siki totobo ” a thing ”, pidgin English “ one thing ” ; siki hava ” some-
thing ” ; siki nggari mane “ a boy ”. See § 53 for ki and siki.
The article na is used with nouns which have the noun termination
in a, see § 11, and also with words which have no definite noun form
as such : mate “ to die ” ; na tnateadira ” their death ” ; bosa " to
speak ” ; na bosa “ speech ”. It is used also with the ordinal numbers ;
na ruani, na ruana “ second ” ; and with hangalatu “ hundred ”,
toga “ thousand ”, mola “ ten thousand ”.
4. There is a second article na which denotes ” belonging to ” a
place, like the Bugotu gna and the Vaturanga na ; na Belaga ” a man
of Belaga ” ; na Galilea igoe “ y'ou are a man of Galilee ”. Mr. S. H.
Ray connects this na with the ligative article na in the Indonesian
languages.
There is a use of this na in both Florida and Bugotu languages
to denote places : na Betidalo “ Betidalo ”, the home of the dead ;
na Babylon “ Babylon ” ; na Boli “ Boli ”, a village.
5. The ordinary article na preceded by ra, the personal pronoun,
3rd pers. plur., “ they,” is used to denote the plural of persons : ra
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1079
na huh “ fools ” ; ra na hogona “ his relatives ” ; rana tinoni “ men ” ;
ra na tamahogoni “ kinsmen ” ; see § 17.
There is an idiomatic use in the separation of the pronoun ra from
the article na : ngge ra mai ta a Jesus na vaovarongo then came the
disciples to Jesus ” ; e gua ngge ra bosa na tarai “ why do the Scribes
say ? ” ra in these phrases is evidently used as a subject, and there is
no further indication of a plural. The ordinary plural article ra na is
used when tara, Jcara precede with the verb and the subject follows.
A plural is also denoted by the suffixing of ra, the pronoun of the
object, 3rd pers. pL, to verbs or certain prepositions, or of dira, the
pronoun of possession, 3rd pers. pi., to nouns or certain prepositions,
the article na following with the noun in both cases without any further
plural sign ; tu rigira na tinoni “ 1 see (saw) the men ” ; tara nira
belebele na Heburu “ they disputed with the Hebrews ” ; na halautudira
na sinogo “ the path of the strangers ” ; tadira na tinoni “ among
men ” ; ma ra vetenara hei “ and they sent certain people ”. This is
the use also with didia, didira “ their ”, and nggaira “ they ” ; didira
na vugo “ their nets ” ; nggaira na mane “ the men ”.
When plurality has been already expressed by the suffixing of the
plural articles gi, i (used of things), to the verb as anticipatory objects,
the noun of the object is used with the singular article na : kau
vavonugi na popo “ fill up the pots ; te bosai vania na bosa eni “ he
spake these words to him ”.
A noun of assemblage, lei, is used preceding the noim to denote the
plural both of persons and of things : na may precede : na lei tinoni
“ men ” ; na lei dahna his sons '' ; the possessive noun ni, with a
suffixed pronoun, may intervene between ra and na : ra nimua na
lei nggari “ thy children ” ; hi may be used without na to denote the
plural: hi tinoni “men”; hi totobo “things". This hi occurs in
Sa‘a as ahi and in Ulawa as alai, ala, all used of persons only. Bugotu
has komi, koi, used as nouns of assemblage, but of persons only.
In Psalm Ixviii, 12, the phrase na koi hanu “ persons ” occurs in Florida,
but its use is doubtful.
6. The personal pronouns ara, ra, are used by themselves with the
cardinal numbers to denote a specific set of persons : ara kangavidu
“ the ten ” ; ra rua “ the pair ’’ ; ara rua na tinoni “ the two men ” ;
ra vitu “ the seven " ; nggaira ra lima “ the five ” ; ra na tamahogoni
ara vitu ngga “ there were seven brethren " ; ara ngiha na mane
“ how many men are there altogether ? "
7. The personal article a is used with aU personal names, male and
female ; a Manoga, “ Manoga ” ; a may be used to personify : a tupi
vatu “ the stonemason ” ; it is used with the relationship terms to
1080
W. G. IVEXS —
denote a specific person : a Dale " the Son ; a tinana " his mother ;
also with hanu ' person ; a hanu " he who the person who .
so-and-so ; it is also used with the numerals siki, saJcai “ one ;
a sakai one ” ; a siki sakai " some person " ; a siki tinoni " a certain
person ; a sakai na havgnlatu “ one hundred ” ; a sakai vai/iun
only once After the usual custom in Melanesia, the personal
article is not used with vocativ^es.
III. Nouns
8. Names of parts of the body, the relationship terms (except
the vocative mama “father’), and words denoting position : near,
opposite, above, below, take the suffixed pronouns of possession ng(]u,
mu, na, etc. However, nearly all nouns take these suffixed pronouns ;
but the words for “ clan ’ kema, “enemy” kana, leva ni mate, “neigh-
bour ” koynu kolu, “companion” udu, “belt” kabe, “shield” tako,
are used with the possessive noun ga, the pronouns of possession being
suffixed to it ; na gana udu ” his companion “ ; na gamiu na kana
jour enemies ; see § 29. Some nouns may be used with the
possessive noun ni with the pronouns suffi.xed as well as taking the
suffixed pronouns ; nina bosa, na bosana, “ his speech “ ; nina eale,
na lalena his house . It does not .seem that the use of nina, etc.,
denotes emphasis, for the uses are interchangeable.
9. Any word which in form is a verb may be used as a noun :
bosa to speak , na hosa " speech ’. A verb with its adverb is treated
as a noun : bosa tabo to bla.spheme ", na bosa tabo “ blasphemy ' .
A noun form inaj^ denote purpose : na vano “ to go ”, na nggito " to
steal ; this form, which is gerundival, is used after the verb
tangomana “ to be able ; see § 3.
10. There is an idiom in Florida which is not found in the kindred
languages of Bugotu and ’\ atiiranga, viz. a verb with its verbal particle
ie or ke is used as a noun, “answering,'’ as Codrington savs, " to a
participle in Greek, or to a pronoun with a relative clause in English " :
ngge bosa vania te rapobde " then said he to the paralvtic ” ; te mua
leiu ni pala lanigita, te nigita udu " he that is not against us is with us ’' ;
na hrnana ke pernu “ the hand of him that betrays me ”.
11. There are two noun suffixes employed in Florida -n. ni,
but neither has a \er} vide u.se ; a is added to both verbs and nouns ;
jatc “ to die '', matea “ death ” ; kutn “ to fall ”. kutua “ fall ” ;
ato to go up daloa “ ascen.sion These are only used with the
suftixed pronouns of possession.
A GRAMMAR OP THE LANGUAGE OF FLORIDA
1081
The phrases nggariamu ” in thy youth nggari maneanggu ‘‘ in
my youth occur in the texts, showing the suffixing of a to a noun.
There is always the danger of confusing this noun-suffix a, used with
verbs, with the gerundival particle a ; but the latter is only used with
transitive verbs.
For the noun-suffix ni see § 15.
12. The gerundive. There is a use in Florida of the article na
with a gerundival form consisting of transitive verb, a gerundival
particle a, and the suffixed pronoun of possession in all persons and
in both numbers, and not merelv' in the 3rd pers. sing, and plur. as in
Vaturanga. (The gerundival suffix cigna in Bugotu is probably a com-
pound of a, the gerundival particle used in Florida and Vaturanga,
and gna, which is used as a gerundival suffix in Bugotu).
Codrington regards the Florida gerundival particle a as conveying
a sense both passive and active ; but this particle is only attached to
transitive verbs in all of the three languages (as is also the Sa‘a, Lau,
and Ulawa gerundival particle la), and therefore must always be active :
na bekuanggu ” to bury me ”, my burial ” ; na bekuailira ” to bury
them ” ; the verb may have a second verb (used as an adjective or an
adverb) or a preposition following, in which case the gerundival
particle a, and the pronouns of the object, are added to the second
member of the compound : na labu taboana ” to kill him without
provocation ” ; na bosa vanianggu " to speak to mo ”. The ordinary
personal pronouns may follow a gerundival use, the second pronoun
being used objectively in addition to the suffixed pronoun of the
object : oia tabotaboamiu igau “ to tempt you ”. There is no idea of
the second pronoun being added for the sake of emjjhasis. A more
definite object may follow' : 7ia rongoriana na bosana " to listen to his
words ”. This gerundival particle a is probably the same as the
gerundival particle la of Sa‘a, etc., through the loss of 1.
It is perhaps because w'e are accustomed to say in English ” my
being killed ”, rather than ” the killing of me ” that the Florida
gerundive was classed by Codrington as both passive and active. But
the Melanesian method of expression is the second of these two phrases.
The verb tangomana ” to be able is used in Florida and Bugotu
with the gerundival form : tarn tangomana na kisuana “ they were
able to cure him”. The verb lavi bide “to annoy”, “vex”,
“ trouble ", is used with the gerundival particle a after the Bugotu
fashion, i.e. with a pronominal subject : e gita ngge au lavi buleana
“ why do you trouble her ? ” There is a single instance in the texts of
this use in the case of a second verb ; ma ra sari taoniana nggaia
“ and they followed him ”.
13. Genitive. A genitive relation is showm (1) by the use of the
1082
W. G. IVEXS —
prepositions m, i. According to the regular practice in the Solomon
Island languages, no article is used before the noun which follows the
genitive prepositions. Xi is in common use as a genitive, but i is only
used in certain phrases: na dale i boh " a voung pig " ; a Dale i
tmoni " the Son of Man ” ; na mate i gabu “ the price of blood ; na
bonqi i vavola “ the day of salv'ation
In Florida and Bugotu ni i.s not used with the second of a pair C)f
nouns joined by the copula : na l i ni Betidalo ma na mate " the keys of
Hell and of death ' ; na talugu i sea nia na lapa “ the outside of the
cup and the platter ”,
•A i is used after the verbs vuni to begin ”, lilin to change ”, to
turn into ”, viiha to become In Sa‘a the verbs for “ begin ” and
“ change ” are followed by the genitive ni.
A't denotes “ purpose ”, as in Sa‘a : te sopou ni vanga “ sitting at
meat ” ; Ice kabu ni tunava “ it will be for an inheritance ” ; na vatu
ni piniti “ a stone for anchoring
(2) By the use of the suffixed pronouns of possession, 3rd pers. sing,
and plural ; na tamana na nggan “ the father of the child ” ; ra na
daledira na prophet “ the children of the prophets ".
(3) By the use of the po.s.se.s.sive noun ni, in the case of persons :
na sonihalavn na nina a God “ the grace of God ”.
14. Prefixes to nouns. An instrumental prefi.x is seen in ignho
“ a digging stick ” ; gaho " to dig with a stick ” ; ikara “ a baler ",
kara to bale ?/(nh( “ a needle ", Sa‘a " to boro ”.
The prefix tama is used with certain relationship terms to denote
a company or set of people ; tamahogoni “ relatives ” ; tamavavine
brother and sister ; tnmatahi ” brothers ", sisters " ; taniatau
means husband and wife tan " spouse ” ; tamadale “ father and
son ; ra na or na lei precedes. The prefi.x tama occurs also in Bugotu
and A^aturanga with a similar use.
1j. Suffixes to nouns. See § 1 1 for the noun-suffix a. The suffi-X
ni is added (1) to nouns to denote a body of people : tamahorjoni
relatives , hogo a near relative ’ ; tamavavine, tatnavavineni
brother and sister , varine “ brother or sister ” ; na ova deteni
the Judges ; ra na or na lei precede.s : (2) to the cardinal numbers
to denote the ordinals ; rua ‘‘ two ”, na rnani " the second ” ; na
appears also as a suffix in the case of certain numerals : na ruana the
second ; na hangnvnlnni, na hangavnluna “ the tenth ”.
16. The reduplication of a noun shows inferiority ; tinoni " man ■
inoni image ; rale ‘ house ' , “ shed ” ; “ coconut ”,
A GRAJIMAR OF THE LAXGUAGE OF FLORIDA
1083
niuniu “ a wild palm ; gotu a mountain gotugotu “ a little hill ”,
“ a heap ’ ; manu '' bird ”, maumanu “ insect
17. Plural. Plurality is denoted by the use of (1) the plural articles
ra na of persons only. (2) the noun of a.ssemblage lei u.sed of both
persons and things ; see § 5 ; and (.3) the plur. pronoun, 3rd pers.,
nggaira: nggaira nina mane “ his men'' ; §21.
Codrington states (ML., p, 525) that ra is used alone to denote
plurality, and instance.s ra hogonggii ” iny brothers”, ra hihiadira
" their ancestors ” ; but the texts do not bear this out except when
ra is used with the possessive noun ni, or when ra i.s used of a number of
people, with the numerals : ra ninggaa na lei nggari ; ra vilu : other-
wise na is always added to ra.
The repetition of a noun, with the copula ma, serves to denote
plurality : na varaia ma na rarata “ generations ”. This use is found
also in Bugotu and Vaturanga.
Two plural suffixes, gi, i, are attached to verbs as objects when
things, and not persons, are in question ; see § 22. When the verb
ends in f, gi and not i, is the suffix used : in other cases i is used :
(ara idumigi na vundu ni niumiu “ the hairs of your head are
numbered ” ; tara sitki potai na linianggu “ they pierced my hands ” ;
te vagai “ like ”, of many things.
Codrington’s example (ML., p. 525), lei iga te subo tara hdlai,
shows na lei iga “ fl-shes ”, used as if a singular, te subo many ”, not
tara subo, and the ]tlural suffix i added to Iiola ; this is because na lei
iga is used collectively, and this nractice is found also in Bugotu ; so
also with his other example na lei hongi te mai, where na lei bongi
" nights ” is used collectively with a singular verb, te mat " comes ”.
This is the general use when the noun is neuter, and in the plural.
In Sa‘a ‘i (for gi) is used .similarly to Florida gi, i, as a plural object
of things.
The suffixes gi, i. are used also of ordinary plurals ; e rua na range
te pile hhgi ” two very small shell-moneys ’’ ; utoi tua “ very good ”,
“ that will do ”, of many objects ; anggai " this ”, anggai gi ” these ”,
A plural noun suffix ni is used of 3rd pers. plur. (of things) with
certain nouns instead of dira: vula "moon", "month”, vulani
" their season " ; ninlu " year ", ninluni “ their seasons " ; na snleni
" the big ones " ; na pileni “ the small ones ". There is a similar use
of ni in Bugotu, Sa‘a, and Ulawa,
Totality is expressed by udolii (Mota nol). and completion by soko :
na hulimu udolu "thy whole body”; m udolu '‘the whole”, “a
1084
W. G. IVEXS —
round thing ", " a loaf ; na Jei vavata soJco “ all the generations ;
soJio tua " finished ” “ that will do ! ; c ngilia soko how many in
ain ” “ as many soever
As Coclrington states, lei, a noun of assemblage, conveys by itself
the notion of totality ; ninggua na ki totoho ’’ all that I possess
18. Gender. To denote gender mane male ", vaivine ■' woman ".
are added to the noun : nggari mane " j'oung man ", bov ' ; a dahi
Sion vaivine the daughter of Sion
19. The word tina “ mother is used of thing.s that are big :
na lei tina ni heti maiiri " rivers of living water " ; liala tina " a high-
way ” ; beti water beti tina " river ’’ ; j/ari " earth pari tina
'■ niainland na niuhi tina " high summer . This use is found also
in Bugotu and Sa‘a.
20. Five nouns — hanu, vatei, 7nal€, pukn, pile — deserve notice
Jiaiiu denotes " person, the person who, he, so-and-so ”, the
personal article a preceding : a hanu te mate tun, a Christ sugua " one
died, namely Christ ’ ; a hanu is used when a person’s name is not
remembered. Similarly na hanu means " what is the tiling 1 ”, " what
do you call it ? ”, when the name cannot be recalled. Bugotu uses
ha>iu in the same way, and Codrington quotes the Oba, New Hebrides,
hen, heno {ML., p. 134) as the same words, and equates them all with
the Malagasy heno. In MIL., p. 40(). Ray has an instance ra sanu
havui ■“ new things ”, in a language of the New Hebrides, where sanu
is probably the Florida hanu.
1 atei denotes thing for doing, place of ”, and one may compare
its meaning with that of the Bugotu hali ; it is used with the article
na preceding. Codrington says that vatei is probably the same as the
Mota vatiu ■■ place ”. He treats vatei us ” the object of an action ”, i.c.
as having a pas.sive sense ; his examples are {ML., p. .525) {na) vatei
sukagi “ a thing sacrificed ; {na) vatei aroviamnnii “ the object of the
pitying of us ’, “we pitiable objects ” ; but e.xamples in the texts do
not bear out his meanings : na vatei bita win-e “ a wine-press ” ; «« vatei
inu ‘ a drink ” ; na vatei sosoni " a thing for giving ”, ” tribute
money ” ; «a vatei lala belt " a thing to draw out water with ” ; 7ia
vatei talu hulu “ a thing (place) to put the lamj) on ” ; na vatei hevei lee
an offering ” ; 7ia vatei ke nia hare " a thing for wonder ” ; an lei
vatei tootoro “ tribulations ” ; na vatei na nia haruadira " a thing to
accuse them of ”.
It is evident, then, that (na) vatei sukagi is “ a thing for sacrificing ,
rather than " a thing sacrificed ’ ; while (na) vatei arnviamami means
the pitying of us ”, “ to pity us As with the gerundival particle a,
§ 12, the Melanesian mind .sees things from the active rather than the
passive point of view. The Sa’a leu and the Ulawa lehu mean either
place or ' thing ”, and thus bear a likeness to vatei.
Male is always used with the genitive i attached, and Codrington
quotes it as malei ; it seems to be connected with mala “ place ”
t ugotu matka, Nguna, New Hebrides, malo), and its meaning, with
A GRAilJIAR OF THE LANGUAGE OF FLORIDA
1085
i added, is “ place of ” : na lei malei beku tinoni “ the places of men’s
burial ” ; na malei kabu “ a place of sitting ” ; the article wa precedes.
Piiku means source ”, “ beginning,” " chief,” ” real,” ; the
suffixed pronoun of possession, 3rd pers. sing., 7ia is added : there
is a use also of puku as a verb meaning " to emanate from ”, ” have a
beginning in ”, " perceive,” “ know,” ” understand ” : na puku vaka
" the owner of the ship ” ; na puku koinu, " a householder ’ ; vaa
puku rnai to come as from a source ” ; na puku ni sukngi “ the chief
priest”: na puku 7ii mana “the source of power”, "almighty”;
7ui puku ni kema “ patriarch ” ; nxt pukuna inan " for mv sake, because
of me ” ; na pukuna ke (keri) " therefore ”. The frequent use in the
texts of na pukuna " because ”, is incorrect, being a mere following of
the Mota ape, apen.
Pile is used as noun, verb, adjective ; its meaning is “ a little
“ to be little,” “ little ” ; siki pile “ one piece ” ; siki pilena “ a little
piece of ” ; siki pile hulina “ the least of his bones ” ; siki pile totobo
“ a small thing ” ; na pileda “ our share ” ; pipilena “ a little ” ; na
pileni “ small ones ” ; na lei pilepile vanga “ small fragments of food ” ;
yia nggari pile “ a small boy ” ; inau tu side, igoe to pile “ I am bigger
than you ”.
IV. Pronouns
21. (1) Personal. Pronouns used as the subject of a verb ; —
Sing. 1. inau, nau, u.
2. igoe, o ; goe.
3. anggaia, nggaia.
Plur. 1 inch igita, gita ; a.
1 exel. igami, garni ; igai, gai, ai.
2. iganm, gamii ; igaic, gaii, au ; agau.
3. anggaira. nggaira ; ara, ra.
Dual 1 inch irogita, rogita.
1 exch irogami, rogami ; irogai, rogai.
2. irogamu, rogamu ; rogau.
3. ironggaira, ronggaira, aronggaira.
Trial 1 inch itohigita, tolugita.
1 exch itolugami, tohigami ; itolugai, tolugai.
2. itolugamn. tolugamu ; itohigau, tolugau.
3. itolunggaira, tolunggaira.
The dual and trial forms contain the numerals ro “two”, tolu
“ three ". The trial number i.s used of three persons.
The initial i of the forms in the first column is detachable. The form
yiau, 1st pers. sing., is not in common use : e mua nau Jiegenggu “ I am
not alone ”. The .short forms, ?/. o, and ai, au, are used by themselves
as subjects without being compounded with the verbal particles te.
1086
W. G. IVEXS —
ke : 0 mua ganagana a tahinggu e saniu ngge u nggehenggehe hegenggu
“ do you not know that my sister has left me to serve alone ? ’ e gua
ngge ai mua tangamana “ why were we not able ? ’’ e gua ngge au
matagu why are you afraid ? ai, au may be followed by the longer
forms, gai, garni, gamu : mai hangavia, mi garni tai mua rigia " and
we opened the door, hut we did not see him ; goe " thou ’’ is used
only in address, e goe ! a goe !
The pronouns of the first column in sing, and plur. numbers are
followed by the short forms u, o, a, ai, au, compounded with the verbal
particles te, ke (the vowel of the particle being elided), in the forms
tu, ku, to, ko, ta, etc. ; and the pronouns of the second column in the
plural are followed by the short forms a, ai, au compoimded with te
and ke ; the forms igai, gai, igau, gau are used as subjects, and the
longer forms in each case may be followed by the short forms com-
pounded with te and ke : e gua gau nia hare “ why do ye wonder at
it ? ” igai tai tugunigi “ we declared it ”.
The pronoun a is only used when compounded with te, ke in the
forms ta, ka ; Vaturanga uses it in the uncompounded form ; the
pronouns u, o, ai, au are compounded with the verbal particles te, ke in
the forms tu, to, tai, tau, ku, etc.
There is a certain dislike to using the forms in the 3rd sing, and pb,
anggaia, nggaia, anggaira, nggaira, of things ; anggaia, nggaia both
bear the meaning " he that ”, “ the person who,” “ that which ’ :
they are also used before personal names as imanea, manea are used in
Bugotu, and aia in Vaturanga : anggaia a God ; na maiana nggaia
te maemane the coming of the righteous one ” ; nggaia te kikokiko
“ the deceiver ” ; anggaia is not used when the preceding word ends in
a, nggaia serving instead.
The pronouns inau, igoe, igita, igamu, igau, nggaira are used as a
kind of secondary object after the verbs and prepositions to which a
pronominal object is already attached ; itamiu igau " from, to,
wth, by, you ” ; see § 12. The initial vowels of inau, igita, igarni,
igamu, and aronggaira are displaced by mua " not ” and sopa
“ each " : i mua nau ; i sopa gita.
The pronouns igau, gau, agau are used as plural vocatives, and
agau has no other u.se : o (ko) serves as singular imperative with a verb.
V, gita, gai, ai, gau, ra are suffixed as objects to verbs and
prepositions ; when used as subject, igami may be followed by garni,
and igamu by gamu, and gau by au ; the compound forms tai, kai,
tau, kau always follow igami (igai) and igamu.
Ara and ra are used with the numerals, see § 6 ; ara is a compound
of a, personal article, and ra ; the personal article a appears also in
agau, aronggaira ; for the use of ra to denote the plural, see § 5.
Ra is used by itself as .subject : ivei ra tara hurugo ’’ where are your
accusers ? ” ra vulitabua they baptized him ” ; ra a, ra gagua “ they
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1087
Ra is used to introduce the subject, as it were, when the verb precedes
the subject, see § 5 : ma ra vaukolii rnai na vure subo “ and a great
multitude was gathered together Ra is used with the verbal suffixes
te, he in the forms ta ra, ka ra ; these are usually written in one. For
ta ra, ka ra, see § 1.
Ara and ra are used with liei “ who ? ’’ " any one ’’ ; araJiei, rahei
who ? ” " they who,” “ whosoever ” ; ra is used with the demon-
stratives ini, keri, kiri, to denote the plural : ra ini, ara ini “ these ”,
rakeri, rakiri " those ”.
Nggaira is used to denote a plural : nggaira nina vaovarongo “ his
disciples ” ; nggaira na Pharise " the Pharisees ” ; see § 17.
Ro “ two ” is used by itself as meaning " we two ”, ” you two ”,
“ they two ” : e gua ngge ro nea eni " why are you two doing this 1
and is added to the particles te, ke, the vowels of which change to o :
irogami koro tangomana we can " ; ronggaira toro bosa “ they two
said ” ; koro tono “ you two go ”.
Toluganii, tolugai are followed by the form tai tola, and tolugamu by
tau tolu, etc. : tolugamu tau tolu kenea “ you three searched for him ” ;
tolunggaira tara tolu iona ” they three went ”.
22. (2) Pronouns suffi.xed as objects to verbs, and to prepositions
which are verbs in form : —
Sing. 1. u. PL
2. go.
3. a. 3. ra,gi,i.
The forms gi, i are used of things only, see § 17.
As stated above, § 21, the longer forms of the pronouns, inau, etc.,
are added as objects in addition to the suffi.xed pronouns. This is not
done for emphasis, but is the regular use. The pronoun of the 3rd pers.
sing., a. is used as an anticipatory object, following a verb or a
preposition ; there seems to be some latitude in this use of a with
regard to verbs, i.e. it is not always suffixed to verbs, but it is always
added in the case of prepositions ; rania a Lord " to (him) the Lord ”.
However, when a transitive verb (i.e, one to which the pronouns of
the object can be suffixed) is followed by' another verb, or by a verb
used adverbially, the pronoun a of the object is always added to the
second verb as an anticipatory' object in agreement w'ith the
pronominal object : kara vadangitaili leea nggaia “ they will persecute
him ” ; te sani kasilau inau " Icav’cs me for good ”. In a similar
way ra is suffixed in the case of the plural.
In the plural and dual the personal pronouns gita, garni, gai, gamu,
gau are suffixed as objects ; in the dual and trial numbers the forms
without initial i or a are thus suffixed : the numerals ro. tolu precede :
ke nea ro vanigau " will do it to y'ou two ". In the dual and trial
1088
W. G. IVENS —
numbers the forms of the personal pronouns without initial a follow
the suffixed pronorms of the same person : te bosa mnira tolanggaira
spoke to the three of them
Certain verbs have the pronouns of the object attached in all persons
in the singular and plural numbers, in what Codrington calls " a middle
use ' ; gaha, gana ” to dwell ” ; kode " to be in vain ” ; taga " to be
lost ” ; vaga " to be like are thus used : tu mun sama me kodeu " I
have not run in vain ’ ; tara gahara " they dwelt ” ; tarn tagara " they
were lost The verb vaga " to be like " has two uses ; the suffixed
pronominal object agrees (1) with the thing (or person) with which the
comparison is made : (2) with the person making the comparison, or
about whom a comparison is made, this being an example of a " middle
use : (1) te vagara na sheep, " like sheep ’ ; ivei te vayaa na tinoni eni
“ what is this person like ? " “ what sort of person is this ? ” na 1^*
totobo vagai rami " things like these ’’ ; kau bei te vagara " be ye not
like them ” ; te kenera te vagara he seeks such as they ” ; (2) inau tu
vuha te vagau na Judea '' I became as a Jew ” ; te vagagau na tinoni
pepelu “ you are like a merchant ” : ma kau vagagau nina mane ni lutu
" and you shall be like his servants
Codrington's example {ML., p. 535). te vagngai na dale ni bolo te tona
sania na tinaiia we are like a young pig whose mother has left us ,
is hardly patient of his translation " a young pig is like us It is rather
an example of a " middle ” use.
The suffixed pronoun na, § 23, may be <added to vaga, making the
word vagana, which means " very’’ : a Marg vagana eni “ this was
that very llary
There is a double usage al.so in connection with the verb ho ‘ to
desire, wish”; “heart”, “mind”, “will”; (1) inau tu mna liona
“ I do not wish ” ; (2) inau tu nma lionggu “ I do not wish ” ; man
tupono lionggu na ganiana na ranga “ I forgot to eat my food ’. The
second and third of the.se examples show a “ middle ” use.
The Bugotu word hehc “ mind ”, “ will ”, in the compound vnnohehe
■■ to desire ”, “ wish ”, suffixe.s the pronouns of § 23 in agreement with
the subject : ku vanohehenggu " I desired ”. This appears to bo a
“ middle ” use.
23. (3) Pronouns suffixed to nouns to denote possession ; —
)Sing. 1.
PI. 1.
inch
da.
2.
mu.
1.
excl.
mami
3.
na.
2.
mill.
3.
dia,
dira ;
For nouns that do not use these suffixed pronouns see § 8. The forms
dia, dim are interchangeable, dira being in more common use : ahadia,
A GEAilMAE OF THE LANGUAGE OP FLOEIDA
1089
ahadira “their names’"; valedia. valedim “their houses”: itadia,
itadira ; na bosa vaniadira “to speak to them”; but dira and not
dia is always used when an article and noun follow. Ni is used in the
case of things ; sec § 17.
These j^ronouns denote “ mv, thy etc. ; dalenggu “ my son ;
na niateadira “ their death
To express the dual, ro “ two ” precedes the noun, the plural forms
of the suffixed pronoun being used : na ro limadia " their hands ” ;
iulu ■■ three ” is used similarly for the trial number.
As in Bugotu, da is suffixed to verbs in a kind of imperative use,
where the action proposed is for all those addressed : hageda “ let us
all go up ! ”
Certain verbs have these prepositions suffi.xed as objects : manaha
“ to know by experience ; tala “ to permit ", lio “ to desire ”,
7nana “ to be suited to ”, " to suffice ”, tail “ to marry ". gilala “ to
recognize ” ; of these manaha and gilala also suffi.x the pronouns of
§ 22, and vaga “ to be like " has vagaa, vagana, and vagani. The verb
util “ to be true ", which is usually in the form utuni, has the forms
utumu, utiuniu true for you ! “ in address.
The suffixing of na and dira may convey a genitive idea ; see § 13 ;
and na and dira are also suffixed to the prepo.sitions ta. ita as ordinary
or as anticipatory objects.
The personal pronouns of the first column, inaa, igita, etc., may be
added to nouns (and to the prepositions ta. ita). to which the above
pronominal suffixes have been added. This seems to be quite an
ordinary use, and not done for emphasis : cf. § 22.
Nouns with these pronouns suffixed, and used as prepositions
denoting position, are ; liligi " beside ", liligina “ by the side of" ;
ruvanga “top", vuvungamin “above you"; mra, riiriigii “under-
neath ", sarana “ underneath (it) ".
24. (4) Demonstratives: “this”, "here” aeni, aini. eni, ini,
inei. Ixdtri. hr. keia. anggai ; " that " keri. akcri. ari, kiri : “ these
aini. tni. ini, inci. iri. airi. rairi, raini. nggai gi. anggai gi ; “ those ”
keri. kiri. akeri. akiri. rakcri, rakiri. The forms without initial a or
ra follow the noun or pronoun immediately.
f'odrington expresses a doubt as to whether eni and keri are singular
and ini. kiri are plural. The texts show eni and keri with double uses,
sing, and pi.. “ this, these,” " that, those ” ; both of them being used
following words ending in a vowel other than i or o : when i or o precedes,
the forms are ini, kiri. sing, or pi. These changes are due to euphony.
1090
W. G. IVEXS —
The demonstratives appear to be forms of the personal
pronouns nggaia, anggaia ” he " ; for the gi of cinggai gi see § 17. In
general the demonstratives follow a noun or a pronoun immediately,
but those with the prefixes a or ra may be used alone nggaia eni "he .
" this person ”, " this is he ” ; aeni sugua " he, this one, indeed ;
eni is used with the preposition la, la eni " now ", " to-day.
The demonstratives aeni, aini, airi, akeri, akiri, anggai have a as a
prefix ; this is the personal article a ; aeni means “ this person ,
here ”, ” this very ” ; akeri means " that person ”, " that thing ,
” those ” ; akiri has only a plural use ; anggai is used of things as well
as of per-sons, ‘‘ this,” " here ” : gi is added for the plural and kin may
follow. The forms with initial a may be used at the opening ofa sentence.
Ke is used after eni, ini, keri ; also after ahei " who ? ’ " anyone,
and anggaia " he ”, and after the verbs ganagana " to think ”. gilala to
know ”, soko “ to finish ”, gagua " to speak ” ; also after te vaga " like ,
” as ”, na pukuna " because of ” : te vaga ke " thus ” ; ke gagua ke will
speak thus ” ; anggaia ke " he ”. There is a use of ke to denote a
preterite : ma ra rigia ke " and (when) they .saw him ” ; ta horn tugua ke.
“ which being translated ” ; taho tua ke " never ” ; na aha itamua ke
“ thou hast a name ”. This is the ke of keri, keia, etc.
The ka of kakeri is used to denote a preterite : tu nggerea tua i dania
ka “ which I previously wrote ” ; te sopou me nangi ka " sat and
begged ” ; ka is also used after ahei " who ? ”
A third demonstrative na, wliich does not figure in the above
lists, is used to denote a preterite, or to give emjihasis ; ku huge loka na
“ I will enter in ” ; te taho nina langga na " he has no strength .
Bugotu uses the demonstratives na, hi, ri to denote a preterite.
Iri is explanatory ; anmiu iri " yours ”, “ I mean ” ; anggaira in
“ these people ” ; kakeri means " these ” as well as “ that ”. The
prefix i in ini, inei, iri is the personal article i, just as the prefix a in
aeni is the personal article a : ri enters into the compounds kakeri, ken,
akeri, kiri, akiri, iri, airi, rairi, raker i : airi soko "all these '. The
simple forms of the demonstratives are ka, ke, ni, ri ; the longer forms
are built up from these with the addition of the personal articles a, i.
either singly or together. The simple forms ni, ri do not occur in Florida ;
the former is found in Sa'a and Ulawa, the latter in Bugotu.
It will be seen that in several cases the notion of plurality in the
case of the demonstratives is conferred by the prefixing of the plural
pronoun ra.
Codrington quotes na eni “ this ”, «n keri “ that ”.
25. (5) Tnterrogatives ; ahei “ wlio ? ” whose ? ” ahei didini
na tiola “ who.se canoes are these ? ” ahei na ahana “ who (what) is
his name?" plural rahei, arahei ; hava, ha "what?” "how?
The article nn precedes ; ivei " wdiere ? " " how ? " is used to denote
“ which ? ” ■■ what ? ” : ivei te vagaa " what ? " “ how ? ” also ivia
“ where ? ’ ‘‘ what ? ” " how ? ” The interrogativ'es are also ii.sed a.s
indefinites, “ any one," " whosoever,” “anything," “ whatever."
26. (6) Distributives : sopa “ each ", “ every ”, “ respectively ,
precedes the verb : kara sopa detera “ they will be judged separately " ;
A GRAMMAE OF THE LANGUAGE OF FLORIDA
1091
tara sopa gaegahe “ they severally shouted ” ; sopa is evidently a
verb, since the form sopangi occurs : bosa sopangi “ to speak to each
and all The reduplicated forms are sosopa, sopasopa ; tara vahu
sosopaa “ the only-begotten ” ; the texts show a use of sopa preceding
a noun or pronoun, like the Mota distributive ral, but it is doubtful
how far this use is correct : ta na lei sopa villa " each month ” ; na
lei sopa komu sosopa every village ” ; sopa nggaira “ each of them ” ;
but i sopa gita “ each of us ”, with sopa intervening between the initial
i of the pronoun and the actual pronoun, may well be correct ; sasakai
■’one by one”, “each”; keha, kekeha “another”, ‘‘different”,
” other ” ; na keha “ another ”, ” the rest ” ; keha is apparently a
verb meaning “ to be apart, away, different ” : rahei tara keha
“others”, “some are different”; hui keha “ to remove ” ; pusi
keha “ to cut off” ; talu keha “ to put away”, ■’ forgive " ; kehakeha
denotes “various"; keha is used to denote "one” in a series;
kehani “the first”; balu “some”, with the article na preceding:
na balu “ some ”, “ certain ”, “ a few ”, “ part ”, “ another " ; na balu
gua na tinoyii “ some other people as well ” ; ra na balu “ some people ”;
ki balu “some”, “certain people” ; balu is evidently a noun, and
may be connected with the Sa‘a halu “ some ” ; tango " each ”
precedes the word it qualifies : tango sakai “ to take individually ”,
“ to give to each ”, “ one at a time ” ; te tango sakaira na
nggehenggehena “ each one of them received his work ; tango rua
“ two at a time ” ; tango sakai bongi ‘‘ each day
27. (7) Eelatives. There are no relative pronouns. Their place
is siipphed by a pronoun of the object suffixed to a verb or preposition ;
anggaira tara bungutia tara tutukua ‘‘ they look on him they pierce
him ”, i.e. whom they pierce ; ranira nggaira tara havi mui “ to those
who are still alive ” ; ahei ke mua dolovia a hogona te vaeianea ke " he
that loves not his brother he has seen him ", i.e. whom he has seen.
The idiom noticed above, § 10, conveys a relative sense : anggaia te
nggito ‘‘ he (that) steals ”, “ a thief ”.
28. Eeflexive : hege " self ”, “ of own accord ", ‘‘ alone ”, with the
suffixed pronouns of possession, is used in a reflexive sense : ahei ke
talu sulea hegena ‘‘he that exalteth himself”; mv te inatea hegena
“then he killed himself”; tige "of oneself", "alone", ‘‘of own
accord ”, with the pronouns suffixed, has a similar use. A reflexive
sense is also conveyed by th“ use of pulohi “ to return ”, “ back ”,
following the verb : tau kiko pulohigau " deceiving your own
selves ”.
VOL. VUI. PART 4.
70
1092
W. G. IVKXS —
29. Possessives : ni, ga. These are nouns and are used with the
pronouns of possession suffixed, a being added in 1st and 2nd pers.
sing. ■ ninggua, nhnua, nina ; ganggua, gamua, gana. The plural
forms are dida, nimami, nimiu, didia, didira in the case of m, and
gada, gamami, ganiiu, gadira in the case of ga. The article na may
precede.
The forms dida, didia, didira are irregular, the n of ni apparently
being replaced by d, and it was suggested in the author s \ aturanga
grammar that didira (the Vaturanga form also) might be a reduplica-
tion of dira (see § 23 for the forms), rather than a change from nidira.
as Codrington postulated, there being no known instance in the
Solomon Islands of a change from n to d. Also a further suggestion was
made, viz. that dida (the Vaturanga form also) might be a reduplicated
form of da, the vowel being changed to correspond with the i of gita.
pers. pron. 1st pers. pi. inch These suggestions were founded on the
fact that the forms mami, miu, dira exist in Vaturanga, being used as
suffixed pronouns of possession ; while dida is used like the similar
form in Florida. It may be suggested also that didia of the Florida
forms is a reduplication of dia. It may be that the form didia is due to
the loss of r.
The possessive ni denotes (1) “ my ”, “ mine ", etc. ; (2) " for my
part ”, etc. In the first case the possessives precede, in the second they
follow, the noun : na ro didia (didira) na bosa “ the words of the two
of them ” ; tara tona didira “ they went for their part ".
There is a set of pronouns containing ni and having the personal
article a prefixed ; aninggua, aniniua, anina, adida, animami. antmiit,
adira (adidira), which precede the noun and mean " mine ”, " for Hie
to do ” : animua a God “ thy God ”.
Ga is used of things to eat or drink ; also with the words denoting
“ friend ”, “ enemy ”, “ clan ”, “ neighbour ”, “ belt ”, earth •
“shield”, “sickness”, “ghost”, “death”; the article na may
precede : na gana beti “ his water to drink ” ; na gamua nn vanga
“ your food ” ; na ganggua na leru ni mate “ my enemy ” ; na gadira
napari “ their garden-ground ” ; gadira na vahagi “ their sicknesses ;
gadira na tidalo “ their ancestral ghosts ” ; na gadira na mate " their
deaths ” ; ro na gada na kema “ our two clansmen " ; gadira nn taka
“ their shield ” ; see § 8.
Codrington quotes a further form, ke ; it appears to be used like
ga, but is not in common use.
V. Adjectives
30. The adjectives follow the noun ; and, in general, all words
used as adjectives are in a verbal form, i.e. they are used with the
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1093
verbal particles te, he ; they thus form a predicate. Certain words,
however, are used as pure adjectives, i.e. they are not used necessarily
with a verbal particle, e.g. uto good ", diha " bad ”, side “ big ”,
pile little ” : na nggari pile “ a small child ; na manga side " a
great voice ” ; na parako vaohi ma na pari vaolu " a new heaven and
a new earth ” ; na mane uto " a good man Stress is laid on the
adjectival character of a word by the use of the verbal particles te, ke,
the forms tara, kara being used when the sense is plural.
31. (1) Adjectival suffixes : ga, a ; ga is added to nouns and verbs ;
its use with verbs is the more extensive and may be considered as one
of the most characteristic features of the Florida language. It is added
also to foreign words : leven “ leaven ", leveniga " containing leaven ;
halautu " road ", halautuga " having roads " ; beti tinaga " possessing
rivers”; kuliga "having ears"; gotuga "hilly"; tarungaga
"spiritual”; koru “to go down, be deep", koruga “deep”;
dato “ to go up ”, datoga " high " ; vanga " food ". vangaga " fruitful " ;
rua “ two ”, ruaga “ second ”.
A is added to verbs and nouns, but more frequently to verbs ;
kai, haia " always " ; kiki, kikia ” little " ; uto “ to be good " :
utoutoa “ well " ; macmanc “ to be straight ", maeimnea ” straight " ;
oto “a branch", otootoa "with branches". This a is probably ga
through the loss of g.
In the texts there is a tendency to u.se the termination ga as an
ending of the verbal noun : na vooga “ ignorance ” ; ra na pnkuga
“ the wise ” ; na tabuga " hoIines.s " ; na ovu ni kulaga " a company
of friends ” ; na vatei liuUoga " pleasures ". This appears to be a
copying of the noun ending ra in Mota (from ivhich the trans-
lations ivere made) ; but the u.sage may conceivably be established and
become permanent.
As in Vaturanga, there is a sporadic use of verbs as adjectives, the
suffix li being added : ngasili " firm ". mamavali " heavy ". There are
.several instances in the texts of an ending ha which is sometimes verbal
and sometimes adjectival or nominal : ngarangaraha " to cry ” (Sa‘a
ngara) ; kauha “ firm ", kau " to be fixed ". " fast " ; paupanggnha
“ joints ", panggu " to join ". This may be an equivalent oi ga. since
ha is the equivalent in Vaturanga of the Florida ga (the adjectival
ending quoted above) and ha is used as a verbal noun ending in
Vaturanga.
(2) Adjectival prefixes : ka, ma. ta, tata, tapa. tava ; fhe adjectival
prefix ka is not in common use. and the only instance in the text is
1094
W. G. IVENS —
Muba “ loosed ”, liiba “ to loose Bugotu texts also furmsh a single
instance, kanioto broken ” (Mota niot, Sa‘a rtiou), though there is no
word moto “ to break ” in Bugotu. The Vaturanga texts also furnish
a single instance, kapili “ to roll over ”, pili “ to turn In ML..
p. 169, Codrington refers to this prefix ka, and on p. 187 he states that
it is found only in Fiji ; hut the three instances above, though
isolated, are examples of the adjectival prefix ka in those three
languages.
In MIL., p. 382 (49), there occurs the word kalili “ round about ’ ,
“ round ”, in a language of the New Hebrides. This kalili is evidently
the Florida and Bugotu kolili, with a similar meaning, lili “to go
about ”, a common Melanesian word, and occurring in Bugotu. The
ka of kalili will then be an adjectival prefix, altered to ko in Florida
and Bugotu.
It is more than probable that ‘a, the Sa‘a and Ulawa adjectival
prefix, is for ka and not for^a, as stated in the author’s Sa‘a dictionary .
kali “ to break off”, ‘ahali “ broken off” ; hoh “ to cross ”, ‘aholo
“ across ”, etc.
The prefix is seen in manggoti “ broken ”, nggoti “ to break ;
ta denotes condition or spontaneity ; taboha “ burst ”, boha “ to burst ;
taviikesi “open”, vuke “to open”; tata is a reduplication ot ta ,
tatahaliu “ continually ”, “ straight on ” ; tapa is seen in tapataligu
" to go round about ” ; tapatiiguru “ to rise up ”, “ stand up ”, with
a notion of spontaneity ; tava denotes spontaneity ; tavatogi “ to fall
off spontaneously ”, “ to be loosed ”.
32. Comparison of adjectives. A simple statement may convey
an idea of comparison : anggaia te side, igoe to pile “ he is bigger than
you”; the preposition ta, ita “from”, is used in comparisons: a
Manoga te side ta a Laukona “ Manoga is bigger than Laukona ’ ;
vide “ to be great " is also used in comparisons : migamu tau niua
haba vulera “ are you not much greater than they ? ” The verb vaa
“ to go ” is also used in comparisons : te lada vaa “ more glorious ’ ;
te haba vaa tatana va vatiga na vola “ the life is more than meat ’ ;
tara subo vaa tadira tara diki “ they were more in number than the
first " ; see § 40.
The adverb ngangata “ very ” is used of degree and expresses a
superlative ; it follows the verb ; inau tii tangitangi sale ngangata “ I
cried very much ”. The adjective kikia “ small ” is added to pile
“ httle ”, “small”, to increase its force: pile kikia “too small”,
“ very small ”, “ least ”.
A GRAMMAR OF THE LANGUAGE OF FLORIOA
1095
VII. Verbs
33. Verbal particles. The verb in Florida is conjugated by means
of (1) verbal particles used simply ; (2) verbal particles compounded
with pronouns ; any word used with either of these two methods of
conjugation is a verb. The verbal particles, simple or compounded,
precede the verb ; and in the simple form they may be used without
a subject.
The particles in use are e, te, ke, of which e is without temporal
signification, while te is used of present or general time, and ke is used
of the future or in conditional sentences. The particles precede the
verb immediately.
The particle e is employed of 3rd pers. sing, only, and is not used
with a subject, but is used impersonally ; e uto, me uto “ it is good ” ;
e ntuni “ it is true ” ; e taho “ it is not “ no ”, “ nothing ” ; e wua
“ it is not ” ; e soko “ it is finished ” ; e veujaa “ it is like ” ; e gua
“ how ? ” it is used in the phrases e gagua, e a “ saying ”, of reported
speech ; also it is used with all the numerals from two to ten :
e nia “ two ” ; e ngiha “ how many ? ” The copula tm “ and ” is
compounded with the verbal particle e in the form me, and is used
thus to carry on the tense significanpe even when the tense of the
preceding verb is past, the subject not being further expressed ;
ma nggaia te rugu horu me sari taonia “ and he went out and followed
him ” ; ke vaa me vaa “ it will go on and goes ”, i.e. for ever. Bugotu
has a similar usage, though it was not recorded in the author’s Bugotu
grammar. For this use see ML., p. 173.
The particle te may be used without a subject, (1) in the phrases
te a, te gagua “ saying ”, used of reported speech, te vaga “ like ”,
“ as ” ; also in the idiom referred to above, § 10 ; (2) with adjectives ;
te svbo “ many ” ; and with the negatives te taho, te mua “ not ” ;
otherwise it is used with a singular subject in the ordinary way :
ma nggaia te rugu horu “ and he went out ” ; it is also used in the case
of things : na lei beti te obo pungusia “ the waters overwhelmed it ” ;
see § 17.
Tlie particle ke is used without a subject, (1) in the phrases ke a,
ke gagua “ saying ”, of the future ; ke gua “ how I’’ ke gi(a niga
“ when ? ” ke vitu, etc., of munbers, in a conditional use ; ke vaa me
vaa “ for ever ” ; ke taho “ it will not be " ; otherwise it is used with
a singular subject in the ordinary way : ma nggaia ke hulira “ and he
shall lead them ” ; ke ia used to denote the imperative, and also in
1096
W. G. IVEXS —
conditional sentences ; see §§ 34, 35, It is also used in cases where the
present tense would be employed in English : na hnutna Ae peroii
“ the hand of him that betrays me ; ahei ke dolovia a Jiogona " he
that loveth his brother
Both te and ke are compounded with the short forms of the personal
pronouns u, o, a, ai, au in the forms tu. to. ta. tai, tau, ku, ko. etc.,
the vowel of the particle being elided. The forms in the 3rd pers.
plural are ta ra, ka ni. All of these compounded forms may be used
alone as the subject without a further pronoun preceding ; but
ordinarily the longer forms of the pronouns are used preceding the
compounded forms : gita ka hangga taga igita ” we are about to
perish The particles te and ke are used together, te preceding and
merely strengthening ke, and the sense is future or conditional ; also
the compounded forms are used together, tu ku. to ko. ta ka. ta kai.
ta kau, with ta kara, to koro as the forms for 3rd pers. plur. and dual,
the sense being as above.
This use of the two particles together is an ordinary use. and is not
done for emphasis ; it occurs more frequently in conditional
sentences, etc.
It will be seen from the aBove that Codrington’s treatment of
tu, ku, ta, ka, etc., as mere verbal particles will not .stand, and that it i.s
incorrect to speak of “ vowel changes of the verbal particles m
Florida. The Florida verbal particles belong rather to the third list
of verbal particles in ML., p. 172, viz. those variable by reason of
coalescence with the pronoun.
A definite past is denoted by the u.se of tua following the verb
immediately. This tua is the Ulawa iia " vet ”. " still, " the Mota tuai
“ old ", and the Indonesian /men ; utotua" enough ! " " thatwilldo !
soko tua “ finished " ; e tako tua “ never ", ■' not at all ” ; e tolu tua
na bongi “three days already " ; anggaia te huge tua ta na vale " he
has gone into the house ; tua also denotes finality and emphasis,
like the Sa‘a particle 'oto, which is u.scd of a definite past : me tah<>
vaho tua na pun i “ no darkne.ss at all " ; na tidalo tua ' ' it is a ghost ! :
kau mua tangomana tua na huge " you shall not be able to enter ;
inau tu taho tua “ I shall never" ; anggaia tua " hi.s vmry self" ; h/
no bona ni bongi tua keri “ at that very hour ”.
The demonstratives na, ka, ke are u.sed to denote a preterite,
see § 24.
34. Imperative. For the imperative the verb is either used alone
and without a pronominal subject, or else it is used with the pronoun
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1097
of the 2nd pers. sing, or pliir., either singly or compounded with the
verbal particle he : tiujuru, o tiiguru “arise ! ” hau habu uto “ fare-
well ! ko riigu horn " go out ! ” kau tona keka “ depart ye ! ''
35. Conditional. A particle ke denotes likelihood, probability,
supposing, if, possibly ; it occurs at the end of the sentence, and the
verbal particle ke may be used at the beginning of the sentence :
ahei ke dolovia a kitlana ke " whosoever loves his brother ’’ ; the con-
junction ngge also denotes “if ”, and is followed by the verbal particles
ke, te ke, see § 52 ; ngge may precede a conditional sentence, ke being
added at the end.
A simple statement may convey a conditional sense, the verbal
particle ke being used : ke liona “ if he wills ” ; ke talanggu a God “ if
God allows me"; mu ku mua inuria “except I drink it”; the
particle te may precede : te ke vahagi “if he be sick ” ; igamu kau
rongovia na mangana taeni “you. if you hear his voice to-day”.
Bugotu also uses a verbal particle da, with a future meaning, to
denote “ if ”.
36. Dehortative. The dehortative is bei " do not ! ” “ shall not ” ;
ko bei matagu “ do not fear ! “ tna kara bei pukugita “ and they shall
not know us ” ; na mane ke bei tona sania a fauna “ let not a man
leave his wife”; bei also means “lest”, the conjunction nggea
“ haply ” being usually added ; kara bei nigita kabalagi " lest they be
offended in us ” ; ke bei pungi nggea “ lest it be darkened ” ; bei . . .
nggea also conveys the notion “ must not ”.
sania moans “ do not ” : sania soo “ let be ! ” ke sania “ let it not
be so ”. The verb sani means " to leave ”, “ omit ”, “ reject ”.
37. Negatives. The negative used with verbs is mua ; it precedes
the verb immediately. The verbal particles e, te, ke are all used with
mua : e mua gita ta dolovia “ it is not that we love him ” ; inau tu
mua nggaia " I am not he " ; igamu kau mua mai tanggua " ye will
not come to me ” ; mua comes between the articles na, a, and the noun
or pronoun or gerundive ; a mua Lord ; a mua Dalena a David eni ?
na mua malei lima “ not man's handiwork ". There is a similar use of
the negative boi in Bugotu. Mua also divides the separable vowel prefix
from the pronouns : i mua nau ; a mua nggaira.
The negative verb taho “ not to be ”, ” no ", is used with mua to
denote “ must be ‘h " not fail ” : ma nggaira gua ku mua taho na
laviadira mai “ and them also I shall not fail to bring ’’ ; ke mua taho
siki soosodo ke topora " and there was every chance of their running
into danger ”.
1098
W. G. IVEXS —
Mua, like the Mota negative gate, is used as an exclamation,
“ how ! e mua vaolu na tutunggu ni tarai eni ‘‘ what a new teaching
is this ! " na mua utoutoni vaho na tuadira “ how beautiful are their
feet ! ’’ This use is not found in the Bugotu and Vaturanga languages.
38. Illative. The illative is ngge ; see § 52.
39. Verbal prefixes. The causative prefix is ra, which is used with
verbs and means “ to make ”, ” cause to be ” ; the use of va makes
transitive an intransitive verb : marara “ to shine ", vamarara to
cause to shine ", “ to make light It is also prefixed to verbs that are
already transitive in form ; valadavagini “ to proclaim”, “ extol :
vaguruvagini ” to dispute with ”.
There is a use of the causative prefix va wdth numerals, denoting
times ; vavitu “ seven times "k
40. Auxiliary verbs. The verb vaa is used as an auxiliary. In
itself vaa means “ to go ” ; na kamu tara vaa ngaa the village where
they were going ” ; vaa rigi “ to go and see ”, used as a preposition of
motion towards ; ngge (ke) vaa me vaa “ goes on and on ”, “ for ever ;
This vaa is used (1) before verbs as an auxiliary ; vaa sara “ to go to ;
vaa tova "to go ” ; vaa kage “ to enter ” ; vaa data “ to ascend ;
vaa riti “ to go reach ”, ” to reach ” ; (2) after verbs, continuing the
action, with the meanings “ further ”, “ more ” ; ke mua nia pulugamu
vaa " will he not much more clothe you ? ” ko vetenagami vaa ta na ovv
ni bolo “ .send us away into the herd of pigs ; kau dutuvia vaa a God
“ draw nigh to God ” ; mnrina vaa “ afterwards " ; (3) to denote
comparison, more, rather : anggaia te mana vaa (itatana) “ he is more
powerful (than he) ; te lada vaa “ more renowned ” ; te uto vaa
more blessed ”. There is no similar use of the verb meaning ” to go
in the Bugotu language, but, as wdll be seen in the next paragraph, the
use of vaa in Florida corresponds to that of ba in Vaturanga.
A second word baa, probably meaning “ to go ”, is also used as an
auxiliary, preceding the verb : ahei ke baa gahaa itanggua “ whosoever
abides in me ” ; nuihei te ke baa mai rigigau “ whosoever comes to
you " ; ngge ko baa taidagi ke “ but if thou marry ” ; ngge to ko baa
tangomana " but if thou canst ”. This baa (spelt ba in the texts) is used
also in Vaturanga as an auxiliary verb, and its meaning is “ to go ".
e.g. ke ba me ba " for ever ”, which may be compared with the Florida
ke vaa me vaa.
Codrington (ML., p. 537) treats baa (ha) under “ Conjunctions
and .says it i.s an adverb rather than a conjunction, adding that at times
it must be translated ‘‘ if ” ; but his examples are not consistent with
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1099
the above examples of the use of baa as an auxiliary : (1) ngge kau ha
gilalai na lei totobo ini, ngge kau nia tonggotonggo kau ba kaharigi “if you
know these things you will rejoice if you do them ” ; the first “ if ” is
supplied by the opening word ngge ; the second is contained in the word
kau (see § 35), and ha is used as an auxiliary, the second ngge having an
illative force ; (2) ahei te ke ha hovenggu “ whoever rejects me ”, is a
simple statement, te ke conveying a conditional sense, while ba
strengthens the verb.
Tolu “ to put ”, “ assign ”, is used before verbs as an auxiliary :
talu side “to exalt” ; talu sapa “to embark on a voyage” ; talu
tugu “ to exchange ” ; talu utuni “ to believe ” ; talu tagaa “ to be
lost ” ; talu muri, talu pile “ to be humble The same use of talu
obtains in Vaturanga and Bugotu.
Vuha “ to produce ”, “ bear fruit ”, “ become ”, is also used as an
auxiliary verb following polo “ until ” : polo kau vuha rigia “ until
you see it ” ; vuha is also used as an auxiliary like baa, raa : mi
taeni lau vuha manahana nggaia “ and now you know him ” ; i taeni
to vuha bosa tate “ now thou speakest plainly ” ; «e “ to do “ do to ”,
“ make ”, has an applied meaning : nea na bosa “ to do (i.e. say) a
word ”, “to speak ” ; mivei te nea na gilala “ and how does know ” ?
i.e. “ and how shall he know ? ” See also ni below.
The prepositional verb ni. Certain verbs are preceded by the verb
ni, the pronouns of the object being suffixed ; ni sika “ to hate ” ;
ni gugu, ni tangi “ to cry about ” ; ni maa “ to be ashamed of” ;
nigo hugu “to deny thee” ; nia hunt “to accuse him of” ; nira
vatu “to name them”; nia lunggu “to be tired with it”; nia
tangomana “ to be able to do it ” ; ni tarai “ to talk about ” ; ni vetena
“ to give a command concerning ” ; the last two verbs are transitive,
but the others are intransitive, and ni supplies the transitive force.
This usage occurs also in Bugotu.
It is probably this same ni which is used as a preposition of accom-
paniment, “with” : ni vaa “to go with”, nira vaa “to go with
them ” ; ngge nigi mai na rango “ then he brought the money ” ;
kau nigi mate na palumiu “ you will die of your sins ” ; tara nia kokolu
“ they gathered together about it ”.
There is a further use of ni meaning “ to do ”, “ do to ” : nia na
bosa “ to say ” ; na totobo inau tu nigi “ the things which I do ” ;
mara nigi e vati na tunuva “ and they made four portions of them ” :
ivei te nia “ how do ? ” “ how ? ” tu nia na doloviamiu “ I have loved
you ”.
42. Reciprocal prefix. The reciprocal prefix is vei, veivei ; the
1100
W. G. IVEXS —
transitive verbal suffixes gi and i may be added to tbe compound verb.
gi being used when the verb ends in i. The verb in its complete form,
prefix, verb, transitive suffix, is not transitive, and may denote
relative action as well as positive reciprocity ; veidolovigi to love one
another na veidolovigi '' mutual love " ; veiaravi “ to be merciful ,
na reiarovi “ mercy ’ ; veituilurigi “ to tread on one another ’ .
43. The verbs gagita, a : gagua denotes properly “to do ’ .
“ doings hence “ to say ” (cf. Sa'a te “ do ”, “ say ”) ; it is composed
of ga -f- gua, where gua = “ how ? ” Gagua is used of reported speech,
and is of frequent occurrence. It comes at the end of the sentence,
with the verbal particles e, te, ke, or with iara, kara : e gagua “ saying .
“ said he ” ; tara gagua " said they ”. When its meaning is “ say
gagua is not used in the 1st pers. plural ; when it means “ do ” it is
used with all persons : pe ke mua gagua ke " if it were not so ” ; ivei
tara gagua ‘‘ how do ? ” “ by what means ? ” ma na gagua " and the
rest ” “ what not ” ; na lei gagua “ doings ”, “ things ” ; ahei we-
ngge o gagua “ whom makest thou thyself ? ” te gagua (eni) “ thus ,
“ like this ”, “ namely ”, “ that is ”, “ meaning
A means “ do ", “ say ", and is used of reported speech, occurring
at the end of the sentence. It is used freely, but always of the 3rd
pers. sing, or plural only: e a “saying”, “said he”; ivei te a
“ how ? ” “ how will it be ? ” The forms used are ea, tea, kea, tekea.
ra a : ma na hara itamami, ra a " what is that to us ? said they ’ ;
a is probably the ga gagua through the loss of g. Both a and gagua
occur in Bugotu, with similar meanings and usages, but the exception
noted above in the case of gagua does not occur in Bugotu. In
Florida there is no sign of quotation before the words quoted, but
Bugotu uses e a. The Bugotu gua, of quotation, does not occur in
Florida .
44. Verbal suffixes. The suffixes which are added to verbs to
make them tran.sitive are : —
(1) Simple ; gi, hi, i, li, mi, ni, ngi, ri, ni, ti, vi ; la.
vonu " to be full ", voiingi " to fill ”.
tnngi to wail ", tangihi “ to wail over ”.
knpu ■■ to close ’, kapui to close over knpuri “ to shut tight ’ ,
“ make firm ”.
haye " to enter ”, liageli " to go on board a canoe ”.
SOHO ■■ to swallow ”, sionomi " to swallow something ”.
vua to remove ”, suani " to be carried by a current ”.
.sni7( to cross ”, savungi " to add to ”, “ fulfil.”
hina " to shine ”, hinari “ to shine on ”.
A GRAMMAR OF THE LAXGUAGE OF FLORIDA
1101
ramu “ to beat ”, ramusi ‘‘ to beat a person ”, ” flog
luba to cease ”, lubati “ to loose
inu “ to drink ”, inuvi ‘‘ to drink something
The suffixed pronoun of the object is not necessarily added to the
transitive form, but it is always added to any adverb or supplementary
verb which immediately follows. This is the Bugotu usage also.
The suffix la is used by itself, as in Bugotu, to giv^e a transitive
force to a verb ; in both languages the only instance in the texts of
such use is dika “ to be bad ”, dikala " to harm ” ; but the compound
forms in Florida given below prove the use of la as a verbal suffix.
The compound suffix vagini is added to la, making a suffi.x lavagun>
which is always transitive ; gahe " to shout ”, gahelavagini ” to shout
at ” ; kau “ to be fast ”, “ stuck ”, kaidaragini “ to retain ", ” make
fast " ; tate “ to appear ”, tatelavagini ‘‘ to make manifest ".
The verb nggunalavi {nggitna “ to be disturbed in mind ”) shows
the suffix vi added to la.
For another use of the suffix U see § 31 ; na vakagitaili " hardship ",
(vahagi to be ill ” occurs in the texts).
(2) Compound : hagi, hagini ; kagi, kagi»i ; lagi. lagini ; magi,
magini ; ngagini ; ragi, ragini ; sagini ; vagi, vagini.
These suffixes convey a definite transitive force to the verb, and
may be regarded as transitive suffixes proper, i.e. the pronouns of the
object are regularly attached to them ; but the shorter forms (like
the similar forms ha'i, la‘i, etc., in Sa'a) may be used intransitively.
There seems to be no difference in use between the longer and the
shorter forms, each of them has a transitive force, and the meaning
of the verb is the same, whichever suffix is used.
kia to laugh ”, kiahagi. kiahagini ” to laugh at ”.
kana ” to be hostile ”, kanahagi, kanahagini " to be hostile to ”.
havu ■■ to scatter ”, havukagi, hnvukagini “ to scatter '”.
tau “ to marry ”, taulagi, taidagini " to marry
pari " to take an oath ”, paringagini " to revile ".
poK ? “ to plait ”, poipolisagini '' to do from side to side ”.
tala “ to permit ”, talamagi, talamagini " to consent ”, " agree to ”.
tao " to incline ”, taoni ” to follow ”, taoragi " to stoop ", taoragini
” to overturn ”.
tngiiru “ to stand ”, tuguruvagi " to cause to stand ”, tuguruvagini
” to raise up ”.
A suffix gini appears in laravagini to lean against ”. varava “ to
lean ” ; and in suhipagini to sheathe ”, sulupa ” to go into ” ; tai.
taili appear as suffixes : vadangi " to suffer ”, vadangitai, vadangitaili ;
1102
W. G. IVEN'S —
vahagi “ to be sick vahagitaili to be ill of”. The sufhx Jiagi is
used as a participial ending (like ha‘i in Sa‘a) ; dani “ to be daylight .
danihagi “ till daylight ", “to keep awake all night ’ ; bongi " to be
dark bongihagi “ all day ”, “ till nightfall’ .
45. Eeduplication of the verb. The verb is reduplicated in three
wavs: (1) by the omission of the middle consonant : bongi, boibongi :
this is the usual practice with words of two syllables ; (2) by the
doubling of the first syllable: gilala “to know”, gigilala', (3) b\
the doubhng of the whole word : Jiugu, huguhugu ; this is a practice
with words of two syllables.
Some verbs use a twofold method of reduplication : sopou “ to sit .
soposopou (irregular), sosopou ; pile “to be httle ”, pileptle, pipde.
The effect of reduplication is to increase the sense of frequency of the
action or to intensify the meaning.
Codrington considers the first method of reduplication to be the
normal use, but the texts show rather that the second method is the
more common ; this may be due, however, to incorrectness of spelling,
e.g. titili instead of tiitili.
46. Passive. There is no passive form of the verb ; but a passive
sense is conveyed (1) by the useofta and ka (see § 33), i.e. of the personal
pronoun a compounded with the verbal particles te, ke ; the use of
the first person denoting the participation of the audience with the
speaker : mo, ka mua tango-mana na idumiadira “ we shall not be able
to number them”, i.e. they cannot be numbered ; ngge ta ka lanra
nia na bosa uto “ then shall we receive them with blessings , i.e.
they be blessed ; ma na vangana na maemane ka suba horua ta na
mabo “ and the fruit of righteousness is sown in peace ” ; ta bosa
tugua ke “ being translated ” ; (2) by the use of the verbal particles
te, ke without a subject, i.e. by an impersonal use : ke nei kalei “ it will
be fulfilled ” ; me te mavatigami “ and we are oppressed ” ; te talana
vania “ it is given to him ; (3) by the use of tara, kara, the verbal
particles compounded with the personal pronoun ra, i.e. by the employ-
ment of active forms : kara rat ago nia a Kephas “ thou shalt be called
Cephas ” ; te raga tara arovigami “as we have obtained mercy ;
tara nditabua “he was baptized”. The first method is peculiar to
Florida ; the second is found also in Mota, and the third in Sa‘a.
47. The sentence. The subject is usually at the end of the sentence,
though not necessarily so : te hangga hu na ako “ when the sun was
setting ” ; the subject, if plural and in the 3rd person, is often preceded
by the pronoun compounded with the verbal particle : tara koehorua ra
A GEAAIMAK OF THE LANGUAGE OF FLOBIDA
1103
na Magoi “ the wise men mocked him ; tara mua holo aJiani ra na
vure “ the Gentiles do not name them
There is a very great use of the copula ma in Florida (and Bugotu),
the sentences being co-ordinate and not dependent. Equivalents of
“ when ” and “ while ” are thus not required, and there is no proper
equivalent for “ when but the adverb soo carries the idea of
“ while ” : tiggaira soko tara tona ngge nggenggera, ma a Kurenius te
haba “ they all went to be enrolled, and (when) C}Tenius was
governor ” ; na tangi ma na nggiringgiri live, ma kaii rigira ....
“ weeping and gnashing of teeth and (when) you will see . . .” ; me
bosa vavagaha vanira nggaia te rigira “ he spake a parable to them when
he saw them ” ; te ke gahaa mua “ while he still abode ” ; me te niilavi
“ and (when) it was dark” ; ke rorofo “ when it was morning” ;
ma na maiana mai “ and his coming ”, i.e. when he is come ; te koli
“ as he lay ”.
48. The anticipatory object. The pronouns of the object, 3rd
person, a and ra, are used as anticipatory, or extra, objects of verbs
and prepositions : vania a tamana “ to his father " ; vanira m tinoni
“to the men” ; tara vuivunia na nggenggere eni “they began this
writing ”.
As stated in § 21, an anticipatory object is not always employed with
a transitive verb. The use of an anticipatory object is general in the
languages of the Central and South-East Solomon Islands.
VII. Adverbs
49. Time : mua “ yet ”, “ still ”, “ again ”, of repetition ;
follows the verb ; used following mm not ”, to signify “ never ” :
ke mua durakea mua “ shall never be destroyed ” ; goi ” again ”, of
repetition ; precedes the verb, and carries the notion of " change ” ;
it is followed by gua “ also ” ; gua " also ”, “ again ”, of addition ;
follows the verb ; after the customary Melanesian usage gua follows the
second member of a couplet where “ also ” would not be used in English :
na balu . . . na balu gua “ some . . . others ” ; diki “ first ’ precedes
the verb ; tua “ already ” follows the verb, and is used as a preterite ;
polo, riti, ritini “ until ” ; these are verbs, and the verbal particles te,
A-e follow ; vaho “ then ”, ” thereupon ”, " indeed ” ; has an e.xplanatory
use, or gives emphasis ; precedes or follows the verb ; no verbal
particle accompanies vaho, but ngge may precede or follow, and no
pronominal subject is used necessarily ; taeni, i taeni " now ”, “ to-
day ” ; ngge vaa, ngge gua “ to-day ”, of time to come ; ke dani,
daidani “ by day ” ; ke bongi, boibongi “ by night ” ; i dania, ni dania
“ formerly ” ; sakeragi “ to rise up ” is used meaning ” then, straight-
way ”, and serves as a connective ; soo " if ”, “ whether ” (Lau so) :
1104
W. G. IVENS —
soo inau. soo nggaira “ whether I, or they ” ; nggea “ thereby ” “ there-
with ■■ withal " ; follows the dehortative bei ; also denotes ‘‘ haply ,
" perhaps ” ; soo nggea " haply ", perhaps ’’ ; ta na hule, lakamun
“ haply ” ; i uola " yesterday " ; i kise, ta na ngiha “ bye and bye :
i ngiha when 1 ” " at some future time ”, " bye and bye ” ; vahhn
" the third day on ”, " day after to-morrow ” ; i ropo " to morrow :
i ropo valiha " the third day on ”, " bye and bye ” ; ropo gana. puipungt
gana “ early in the morning ” ; roropo “ morning ” ; vu ni tolu " the
third day on ; vu ni vati “ the fourth day ” (Marau Sound,
Guadalcanal, pui " day ") ; fe bongi hau " midnight ” ; haulagi " of
old ”, ■■ formerly han ” to be far off ” ; tanggiha ■’ recent ”,
Place ; i ani. i nei here ” ; i ane “ there ” ; evei " where 1
“ what ? ” “ wherever ", anywhere ” ; ivei where ? ” “ how ^ ” " by
what means?” "what?” "anywhere”, "somewhere"; 'I'ia
■■ where ? ” " what ? ” " how ? ” " somewhere ” ; kakeri, i kakena
“ there ” ; i knkon “ up ”, " above ”, " on top ” ; i sara, i pari " out-
side ”, " down ” ; i longa “ landwards (from sea) ”, “ north ” ; Ian
" shorewards (from land) ”, “ south ’’ ; the use of lomja, lau as meaning
■■ north ” or " south ” necessarily changes according to whether the
speakers live on the east or the west coast of the island ; boko “ west ”,
" of sunset ” ; uH " above ” ; ulu “ south ” ; galaga up ”, " east :
mai " hither ” ; nggatu " away ” ; dato, huge “ up ” ; horn " down ” ;
ngga, i ngga there ”, " where ”, " thence ”, " there is ”, " there was ,
“ thereby ”, “ thus ” ; follows the verb.
Manner ; lee " merely ”, “ entirely ”, “ just ”, “ very ” ;
“ only ”, ■■ merely ” ; soo continues the action and means “ keep on
doing ”, “ still ”, ” further ”, ” while ”, “ only”, “ merely” ; it follows the
word which it qualifies : inau soo “ it is only I ” ; ma tara vanga soo
" while they were eating ” ; vaasoo “ .still ”, " only ”, entirely ” ; used
of conditional affirmation “ granted that ” ; follows the verb ; vaga,
te vaga ” like ” ; see § 22 ; te vagai ini " thus ” ; ivia te vaga, ivei te
vaga " how ? ” e gua (ngge) “ how ? " ” why ? ” o gim ” how is it that
you ? ” ” how are you ? ” soko “ finished ” ; mate “ to be dead ” is
used as meaning “ quite ”, “ very ” ; beta mate “ quite quiet ” ; ato
mate “ very good ” ; e ngiha " how many ? ” also indefinite ; e ngiha
soko “ how many soever ” ; tovongo " haphazard ", “ to no purpose ,
“ waste ”, “ lightly ” ; precedes the verb ; kulu “ together ”, " with ;
kolu, sonikolu, sakai sonikohi “ together ”.
50. Negative. The negative adverb is taho “ not to be ”, ” no ;
taho tua “ certainly not ” ; taho tua akeri " not that ! ” “ certainly
not ! ” ke taho, te ke taho “ is not ”, *' will not be ” ; see § 37.
VIII. Prepositions
51. Locative i.
Rest at ta, ita, tala, i tata.
Motion to vaa rigi, ta, ita.
Motion from sani, ta.
Dative vani.
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1105
Genitive ni, i.
Instrumental nia.
Accompaniment ni.
The locative i is used with place-names, and also with the adverbs
of time and place ; ta, ita mean " of”, from ", " to (of persons) ”,
“ into ”, “ at ”, “ with (of accompaniment) ” ; the suffixed pronouns of
possession may be added in all persons sing, and pi., except in 3rd
pers. sing., -where tatana, itatana are used ; ta na lei gotu “ to, at,
from, the hills ” ; ta, ita are also used to denote ” by ”, " through ”
of an agent. The forms used with ta, ita are tamjgu, tanggua. tamua,
tatana, tatada, tamami, tamiii, tadia, tadira. ta, ita are used in the
comparison of adjectives, § 32 ; the i of ita is the locative. Ta is not
repeated with the second member of a clause : ta na pui nia na
ungaunga ni mate “ in darkness and in the shadow of death ”.
Vaa rigi means “ go see ”, and denotes ” to ” of motion. It is used
of persons only ; cf. Bugotu regi “ to see ” used as a preposition of
motion towards ; sani is a, verb meaning “ to leave " omit “ not
to do ”, “ reject ” ; when used as a preposition it denotes from ” of
motion ; sania is used also as a dehortative, ‘‘ don't ! " § 36.
Vani is a verb meaning “ go to ", '* say to ", " give to ", " do to " ;
as a preposition it is used either of persons or of animate objects ;
its meaning is “ to ” or for ” ; there is a use cif vani with vctena
“ to send ” ; vetenaa vania “ send him to him ” ; httu vania ” to work
for him ” ; nggaia te nia na bosa vaniadia ” he spoke to them " ; na
vaniana “ to say to him ” ; hea vania “ give it to him ” ; tura vania
“ they said to him ”.
For the genitives ni, i see § 13 ; ai is used of place. " belonging to ” ;
a Joseph ni Arimathea. In Yatiiranga ni is used of “ place whence ”.
Nia means “ with ’’ of the instrument ; it precedes or follows the
word which it qualifies : nia na bcti “ with water " ; le nia pupulua
na hulina na tivi “ clothed with clothing".
For ni of accompaniment see § 41 ; ni followed by hnhi = “ with ”,
to express accompaniment, a verb intervening, pnngisi, piingusi " to
oppose”, “be in the way of” is used as a preposition meaning
“ over”, “ against ”, the pronouns of the object being added.
Codrington's statement {ML. p. 532) that there is no preposition
in the Florida language meaning " concerning ”, '■ in regard to ", is
amply borne out by the evidence of the Xew Testament in the language.
But the translators have not been careful in this matter ; and various
so-called equivalents are found in the texts, e.g. na puhtnn. But in the
1106
W. G. IVENS —
classic cases of John xvi, 8-10, no preposition is used, although in tin’
Mota version from which the translation was made there is a use ot
the preposition ape = “ on accoimt of ”, “ in respect of ” : ke deft
kalea na maramana na tangotatigo dika “ he shall judge the world (ii.
respect of) sin Also in 1 Cor. vii, 1, 25 ; viii, 1, no preposition is used .
ma na lei totobo iri “ now (in respect of) those things
The same lack of a preposition meaning “ concerning ”, “ in respect
of” occurs in the Bugotu language, though in the author’s Bugotu
grammar eigna is quoted with the above meanings. But subsequent
inv’estigations have proved this to be wrong. The translators were
anxious to find an equivalent in both languages for the Mota ape.
Similarly, in neither of the two languages is there a real equivalent
for yap, the “ for ” of argument, though the texts in the languages
show so-called equivalents, viz. na pukuna in Florida, eigna in Bugotu.
These again are only the “ corrupt following ” of the Mota ape. There
are instances in the texts of both languages of the true use, viz.
(1) making a simple statement ; (2) co-ordinating the clause by the use
of the copula ma ; (3) employing the explanatory words mugun.
mm.ua in both languages (see § 54), and hiri, hitagi in Bugotu.
Compound prepositions. These are nouns with the locative
preposition i added: i loka “within”; i sara “underneath”; *
vuvunga “ above ” ; i liligi “ alongside ”, “ beside ”. The pronouns
of possession are added in agreement w'ith the noun.
IX. Conjunctions
52. Copulative : ma, mai, maia.
Disjunctive : ma, pa, hau vaa.
Conditional : ngge, ke.
The vowel of rna changes to i, o, u in sympathy with the first vowel
of the succeeding word ; me is frequently used, even when the following
vowel is i ; ma means “ and ” or “ but ”. The initial i of the pronouns
and the locative i coalesce with the vowel of tna : minau, mi ani.
migoe : ahei ma nggaira “but who are these ? ” ma follows ngge :
ngge me te holo polora “ and he called them aside ” ; the use of ma
is quite a feature of the language, see § 47 : tara ganagana me te nMte
tua “ they thought he wms dead ; nggaia te ganagana ke me te rigia na
rnabubu “ he thought he had seen a vision ” ; ngge ta kau nigi bati na
totobo ini ke, me uto “ if you keep yourselves from these things it will
be well ” ; mai appears to be a transitive verb meaning “ to add to ,
“ in addition ”, and may be the same as Ulawm mai “ to help ” ; the
form maira “ with ” is used of the plural : -maira na tahina a Lord “ with
the brethren of the Lord
As stated in § 33 me carries on the sentence ; in this case the e of
me is the verbal particle ; mai and maia both mean “ and ”.
A GRAMMAR OP THE LANGUAGE OF FLORIDA
1107
The vowel of pa changes to e, i, u in sympathy with the first vowel
of the succeeding word ; pa means “ or The initial i of the pronouns,
and the locative i, coalesce with the vowel of pa ; pinau, pivei : te uto
pe ke dika ; pu tu ku. The forms pe, pi are in common use.
Ngge has an illative and also a conditional and subjunctive use ;
its meanings are “ thereupon “ then”, in narrative ; also " in order
that “ so that ” if ” ; it precedes the verb and is not followed by
the verbal particle te, but by ke, te ke, me ke ; the subject need not be
expressed ; ngge me ke “in order that " ; ngge ke, ngge te ke “if" ;
ngge ke bei “ lest “ ; polo ngge ke “ till ”, “ until ” ; ma a Jesus te
rigia ngge rutu “ and when Jesus saw it he was angry ” ; polo ngge ke
poso “ until it was fulfilled ” ; ngge to ko Dalena a God “if thou art
the Son of God ” ; soko ngge tiva kehaa “ then he put her away ” ;
ngge talana “then he permitted him”; ngge teke mai siki sakai
“if a certain person comes”; when ngge means “if” it may be
followed by ke used at the end of the sentence (see § 35) : ngge inau pe
anggaira ke “ whether it were I or they " \ e gua ngge “ how ? ■’ “ fiow
then 1 ” ngge vaa, ngge gua “ to-day, of time to come ” ; ngge vaa
me vaa “ for ever and ever ” ; nggeni “ to-day ”, is probably for ngge
eni, on the analogy of ta eni “ now ”, “ to-day
The vowel of ngge coalesces with the vowels of the short forms of the
pronouns, nggu, nggo, ngga, nggai, nggau ; ngge au, ngge o, occur in
the texts.
Zfaw “ far from ” is used as meaning “but”, “rather”, “never-
theless ” : me hau vaa ; hau vaa tna.
For ke see § 35. It is not necessary to employ ngge or ke to denote
“if” ; a simple (co-ordinated) statement may be made instead, or
the verbal particle ke may be used : to ko nongi siki totobo, ma a God
ke vahego “if you ask anything, God will give it you” ; ke gahai
rotamiu siki totobo “ if you have an}’thing ” ; migau kau nggehenggehe
uto “ and if you do good ” ; teke mua vahea m-ai i kokou “ if it is not
given him from above. See § 35. See § 49, Time, for soo = “if”.
X. Numerals
53. Cardinals.
1 sakai, siki, ki, siki sakai. 6 ono.
2 Tua. 7 siu.
3 tolu. 8 alu.
4 vati. 9 vilu.
5 linm. 10 hangavulu.
VOL. vm. PART 4.
71
1108
W. G. IVENS
The numerals from two to ten are used with the verbal particle e.
Keha another ”, “ away ”, “ different ”, is used as ” one ■' of a
series {keJceha “ another ”, ‘‘ some ”) ; sakai denotes ‘‘ one, any ■
‘‘a ”, “a certain ; the personal article a is used with sakai and siki :
a sakai vamua “only once”; sakai vamua ‘’one and only one ,
“ it’s all the same ! ” “ no matter ! ” sakai also means “ at all .
“ altogether ”, “ totally ” ; te mua sakai dika “ it's not bad at all :
te sakai kise “ altogether delays ” ; ahe sakai ” to breathe one's last
lio sakai “to dare”, “be brave"; sakai sonikolu “together
sakasakai “ each one ” ; see § 7. For sa of sakai cf. Bugotu sa “ one .
Sesake, New Hebrides, has the form sikai “ one ” (ML., p. 469).
Cf. also Bugotu sikei “ one ”.
Siki denotes “ one ”, “ a ”, “ a single ”, “ some ”, “ any ”, and is
used both by itself and also with sakai : siki vuviilu “ a single hair ;
siki totobo “ a thing ” ; siki nina totobo ” one of his things ” ; siki
■mane “ a certain man ” ; siki nggari iimne “ a certain young man ’ ;
siki sakai “ some one person ” ; a siki sakai “ some person ' ;
siki bona pile gua “ a little while after ” ; siki ranga “ any food .
The ki of siki is probably the ki which is used as an article ; see § 3.
Lau uses si as an article meaning
of a numeral denotes “ each ”,
“ two at a time ” ; vavati ; onoono.
Tee denotes “ only ”, “ single ” ; ganagana sakai tee “ to be of one
mind”; tango sakai tee “to do severally”; sakai vamua na tee
“ single Lau te “ one
Another form of rua is ruka ; Vaturanga also has ruka “ two .
For the numbers over ten, sara “ to reach ” is generally employed .
e hangavulu me sara e rna “ twelve ” ; e hiua hangavulu liiua “ seventy-
seven ” ; na sara sakaini “the eleventh”. A “hundred’' JS
hangalatu ■, a “thousand” is toga] “ten thousand” rnola. The
article na is used with all these ; mola also means “ very many ,
“ countless ’ ; na idu hahi “ difficult to count ” is used as meaning
“ innumerable ” ; inatapono “ closed ’’ may be added after hangavulu
“ ten ” to denote completion.
one " any". The reduplication
‘apiece”, “at a time”: ruarua
Codrington (ML., p. 538) gives a list of specific numerals : banara
“ ten baskets of food ’’ ; goLi “ ten canoes ”, “ ten puddings ;
gaibala ten bunches of bananas ” ; mola “ ten baskets of almond-
nuts ” ; pangga “ ten pigs ”, “ birds ”, “ fish ”, “ opossums ” ; finggu
“ ten coconuts ”, “ breadfruit ”, “ crabs ”, “ shell-fish ”. To these may
be added isa “ ten shell-moneys.” The article na precedes all these.
Bugotu also has pangga “ ten pigs ”, “ a herd ” ; pinggu “ ten
A GRAMMAR OF THE LANGUAGE OF FLORIDA
1109
coco-nuts ” ; and Vaturanga has pinggu “ ten coco-nuts Codrington
also quotes parego “ a ten ”, thing in tens
Codrington gives a set of numerals used in a game at Olevuga,
Florida : ela, ura, lotii, tarn, nila, noa, tivii, rau, beta, taleri ; he states
that these numerals, from the second to the seventh, are metathetic
forms of the ordinary numerals from two to seven, and that eta may be
tea (Mota tea — “ one ”) in the same way. At Sa’a and Ulawa eta is
the ordinary numeral for “ one ” ; rau is possibly for aru, i.e. alu
“ eight ”, by a change from I to r.
(2) Ordinals : ni is added to the cardinal numbers to form the
ordinals ; the article na is used with the ordinals ; rua^ii and ruana
both occur in the texts ; also hangavuluni and hangavuluna " tenth
For “ first ” na diki is used (Bugotu kidi ) ; kehani also occurs : na
diki maladira “their former state”; na diki kema “patriarch”.
The form ruaga denotes “ a second ”, “ another ” : ko bei ruaga na
fupulu “ don't have a second garment” ; toro mua ruaga na vinahi
“ they two are not a different flesh " ; there is a form varuani, meaning
“ second ”.
Multiplicatives. The word tuguru “ to stand ”, “ a post ”, is used
as meaning “ times ” : e tolu na tuguru ni neana ngga “ there were
three t im es the doing of it ”, “ it was done three times " ; sakai
vamua na tuguru ni hageviana “ only once the time of entering " ;
me rua na tuguru ni mate “ twice dead ”. The cardinal numeral may
be used : e tolu na neana '* three its doing “ thrice " ; ke vitu na tango
hahiamu “ seven times the harming of you ” ; vavitu means “ seven
times ”.
Taba “ layer ”, “succession”, is also employed as a multiplicative :
e tolu na taba ni bosa vaniana “ three times the speaking to him
Pape “ together “ in company’’ : lutupape “ to work together ” ;
pape tolu " three together " ; cf. Mota pepe, pepe rua “ tivo together ”,
of canoes sailing.
For “ Distributives ” see § 26.
XL E.xclam-vtions
54. A is used in address : a Simon “ oh. Simon ! " The common
Oceanic e of address is seen in e goe “ you ” (vocative) ; gee, a gee
“ you ” (vocative), Bugotu ge.
Eo denotes assent, teo negation ; ive expresses disapproval ; ina
asks questions.
There are four words which Codrington calls expletives, and which
are used as explanatory words : vamua, mugua, sugua, vagua. Their
1110
A GRAMMAE OF THE LANGFAGE OF FLORIDA
place is probably under “ Adverbs Vamua means “ only ”, “ for-
sooth”, “ I mean”, “ indeed”, “ rather ” ; it is used in explanations,
and is an ecjuivalent for “ because ” ; see § 51 ; it is formed from nnm
“ yet ”, “ still ” ; mugiia, sugna, vagua are all formed from^rMa " also ",
“ again”, of addition. These four words all follow the verb ; mug"''
is used in explanations, like vamua, and means “ rather ”, ‘’I mean
“ don’t you know ! ” It is used as an equivalent for “ because
see § 51 : me mua akeri vamua ‘‘ not that. I mean " ; ma igitu
gua mugua “ and we ourselves also, to be sure ! ” sugua meairs
“ certainly ", “ indeed ", “ to be sure ! ” vagua means “ even ". " ii
it were”, of a supposititious case.
Bugotu has both vamua and mugua, with similar uses.
XII. Oaths, Curses
Codrington, ML., p. 539, gives a number of oaths. To these may
be added a tinada “ by our mother ! "
The Structure of a Bantu Language with special
reference to Swahili, or Form and Function
through Bantu Eyes
By E. 0. Ashton
T he purpose of this article is to bring to notice an aspect of a Bantu
language, which is not often touched upon in the conventional
grammars, and the hope is expressed that its perusal may stimulate
thought. Even if it but blazes a trail from which others may depart,
the writing of it will not have been in vain. This article is both intro-
ductory and complementary to an article entitled “ The ‘ Idea ’
Approach to Swahili ”, which appeared in the Bulletin, Vol. YII,
Part 4, 1935.
As one who has had some years of experience in teaching Swahili,
the writer feels that many of the difficulties met with by a student are
due to a lack of knowledge of the structure of a Bantu language. It
follows, therefore, that the right selection of the main characteristics
of a Bantu language, no less than the jnethod of their presentation,
plays an important part in acquiring “ Bantu eyes " through which to
view Bantu grammar.
WTien beginning the study of any Bantu language, the student soon
learns the importance of affixes, but he does not often realize their
syntactical value, nor that an understanding of their formative value
actually supplies one of the keys to good idiom. In “ The ' Idea ’
Approach to Swahili " already referred to. it is shown that an affix
contains an idea inherent in itself, which it imparts to the root to which
it is affixed ; with the help of a few selected affixes, such as -a. -o, ki-,
vi-. pa-, ku-. mu-, the article shows in detail hoiv these affixes modify
the root. The scope of the present article is wider and more general.
It is : —
(i) To point out what are the principal affixes in Swahili.
(ii) To indicate what the “ idea '' inherent in each affix is.
The application of these principles is left to the reader himself, for
it would require space far beyond the scope of this article to illustrate
each affix in detail ; also, it is in apphdng these principles that he will
acquire Bantu eyes through which to discern form and function. But
1112
E. O. ASHTON
to one who approaches Swahili in the orthodox way, affixes are merely
affixes, which may take the form of Prefix or Suffix, and the work of
Prefixes is to classify Nouns, while that of Suffixes is to make derived
forms of the Verbs, with -ni as a Locative particle. And here, for sucl
a one, the interest in Affixes comes to an end, whereas in reahty e
should be only beginning.
Now the first essential is to recognize that a Bantu language i-
made up of (i) Roots and Stems, (ii) Affixes. With this recognition
there follows the need to examine each component in detail.
Roots and Ste.ms ^
Roots and Stems are either Neutral or Non-neutral in character.
The Non-neutrals are those which can be used to form one part of
speech only ; they are few in number and call for no special comment. -
The majority of Roots and Stems are Neutral in character, that is to
say, by the help of formative affixes and by their contextual situation
they may be used for more than one part of speech. At this stage one
is not prepared to label these roots, in reference to their origin, as
“adjective" roots, “noun” roots, “verb” roots, etc.; that must
wait until more is known of early Bantu speech. Some few roots may
be looked upon as “ Ideophones 'd® In the present article it must suffice
to put them under headings according to their syntactical function m
the sentence. It is the recognition of their neutral character which is
important, for this recognition is the first step towards acquiring
“ Bantu eyes ’’ as mentioned above.
Neutral Roots and Stems
1. Nominal and Verbal. — Mwendo (a journey). Ends, (go).
2. Nominal, Adjectival (or Qualificative) and Adverbial. — uzun
(beauty), mtu mzuri (a fine man), \izuri (nicely).
^ Professor Doke in his Bantu Linguistic Terminology gives the following
definitions : —
Root. — The irreducible element of a word ; the primitive radical form without
prefix, suffix, or other inflexion, and not admitting of analysis.”
Stem. — ” In Bantu that part of a word depicted of all prefixal inflexions.” He goes
on to say : “ Many grammarian.s make little distinction between root and stem.” 'So
distinction is made in this article, and Roots and Stems are classed together, because
the theme is primarily the function of Affixes in relation to both roots and stems.
^ Examples, mimi, wewe, etc.
See Bantu Linguistic Terminology (pp. 11&-19) for some illuminating thoughts on
primitive Bantu speech.
STRUCTURE OF BANTU LANGUAGES
1113
3. Pronominal and Qualijicative . —
(a) Demonstrative -le, li-
(b) Possessive stems -angu, -ako, -akc, etc.
(c) Interrogative -pi, -ni
When these are used as adjuncts to a noun, they are qualificatives.
WTien used without a noun, they are pronominal in function, i e.g. : —
Pronominal usage. Qualijicative usage
(а) Zile si nzrtri. Those are not nice. Xguo zile. Those clothes.
(б) Zile si zangu. Those are not mine. Nguo zangu. My clothes.
(c) Zipi ? Which ones ? Nguo zipi ? Which clothes ?
4. Enumeratives. — -ote.
Lete zote. Bring all of them. Nguo zote. All the clothes.
5. Other Roots and Stems such as -ina, -ekha, -mene, and -mwe
in ChiNyanja.
Affixes
Turning next to Affixes, the ground to be eovered is wider and more
complex. The first division into Prefix and Suffix is simple enough.
(The term ” Prefix ’’ will here include “ Infix ”, as the latter always
precedes the root or stem.) In order to facilitate reference to these
affixes, they are set out in tabular form on p. 1114.
Before chscussing them in detail, it is necessary to point out that it
is not intended that the student should be given a full view of all the
affixes of a Bantu language at one time ; this would put upon him a
burden as rmnecessary as it is undesirable. But it is necessary to dwell
at the outset on the fact that all affixes are formative in function, but
that all formatives have not the same function. The unfolding of their
syntactical function according to their contextual situation must necessarily
be a gradual process. Thus the distinction between Classificatory and
Concordial function in the early stages prepares the way for discussing
later on the difference between hi as a Classificatory or Concordial
prefix and its adverbial function. It follows, therefore, that the
acquisition of " Bantu eyes " is a gradual process. It is, however, a
process which gains momentum in proportion to the willingness to
discard ” European glasses ”, by which is meant the tendency to \’iew
Bantu through the medium of English grammatical terminology and
” word equation ”.
Affixes may be divided into six groups, and each kind of affix has
its own particular function, according to its contextual situation.
1 This terminology does not follow in every respect that of Doke.
1114
E. O. ASHTON —
Prefixes.
m, wa ; m, mi ; ki,
vi ; ji. ma ; n,
n ; u, n ; ku,
pa, mu, with
coalescence.
yu, (m). wa ; u,
(m), i, (mi) ; ki,
vi ; li, ya, (ma) ;
i, (n), zi, (n) ; u,
zi, (n) ; ku, pa,
mu, with coales-
cence.
pa, ku, mu, i,
ki, vi, u
The Initial Vowel
of LuGanda
(and Zulu)
TABLE OP AFFIXES
I. Classificatory in relation to Nouns.
II. Concordial in relation to Adjuncts
of the Noun and to Pronouns.
III. Implicatory of : —
(a) Adverbial Function in relation to
Nouns and Neutral roots used as
Pronouns.
(b) Definiteness to Nouns.
(c) Emphasis on one of two Adjuncts.
(d) Nominal or Pronominal function to
Neutral Roots.
Sufiixes.
-ni
(Classificatory
Prefixes.)
ji
IV. Relationship ; —
(а) Prepositional, to help express an ad-
jectival, adverbial, or associative
idea. (The a root.)
(б) Pronominal to express reference.
(The o root and the c root.)
V. Formative ; —
(а) In relation to Nouns. a, e, i, (ji)- "■
(б) Derogatory idea in relation to Nouns.
(f) In relation to Verbs. ia, ea, ua, l' >'
lea, ika, eka
uka, ya, vya.
za, sha, na.
VI. CoNjUGATiONAL in relation to Verbs,
a, na, li, ta (o) Time.
me, ki, ka, nge, etc. (b) Aspect.
(c) Mood. a, e.
ha, si. [d'j Negation. i
ni, u, (ku), a, (m), (e) Person and Number.
etc., tu, m, wa,
etc.
STRUCTURE OF BANTU LANGUAGES
1115
I. Classificatory Prefixes.
These require no further explanation. Every grammar devotes
space to these prefixes, and deals with them from various aspects,
such as the phonetic laws affecting coalescence and sound change,
the underlying idea in each class, etc., etc.
II. CoNcoRDiAL Prefixes.
These are generally spoken of as The Concords They, too,
receive their full share of attention in all grammars. It suffices to
say here that the concordial prefixes vary in different languages. In
Swahili, for instance, the concordial prefix of the adjective and numeral
is the same as that of the noun, whereas in ChiNyanja they are
different. But this difference is irrelevant to the point under con-
sideration, viz. their concordial nature.
III. Implicatory Affixes.
It is these affixes which deserve more attention than they usually
receive. In form they may be either Suffix, Prefix, or Preprefix. ^
Imphcatory Affixes cover a wide field.
(a) Adverbial Affixes.
(i) pa-, ku-, mil-, i-, -ni. The prefixes pa, ku, and mu, in addition to
being Classificatory and Concordial in function, may in some languages,
such as LuGanda and ChiNyanja, be used as Adverbial Affixes. They
are prefixed either to a Noun (in which case they are Preprefixes)
or to a Neutral Root or Stem to indicate Adverbial function. (In
Swahili -ni, suffixed to the Noun, has the same function as pa, ku, and
mu prefixed to the noun in other languages.) It is probably the three-
fold function of these particles which makes the study of them so
difficult for Europeans. In addition. “ Place ’’ as Subject of a sentence
(which is typical of Bantu speech) finds no counterpart in English
except in the expressions “Here is” or “There is”. Hence the
difficulty in understanding the construction of a sentence such as ;
Pole ahpokufa pokaota maboga ” (= On the spot where he died
there sprang up some pumpkins). In Bantu an Adverb can form the
Subject. In English we should regard “ pumpkins ” as the Subject
in the above sentence.
^ The word " prepretix ” is not used in the above sentence as synonymous with the
■■ Preprefix or Initial Vowel ” of LuGanda. It is used to describe the form of the prefix
in so far as it sometimes precedes a root which already has its classificatory prefix,
that is to say the prefix in so far as it occurs before a noun.
1116
E. O. ASHTOX
The function of these particles, pa, ku, mu is determined by their
contextual situation : —
Classijicatory ; Pahali. = Place (a Noun).
Concordial : Pahali pazuri. A beautiful place. (Concord of
Adjective with Noun.)
Pale pukaota maboga. (Concord of Subject Prefix
with Adverbial Subject.)
Adverbial. Pale pakaota maboga. (Pale — " On the spot ”.)
Place as subject deserves more attention than it is generally
accorded. The use of pa as subject directs the emphasis on to " place "
rather than on to “ pumpkins and illustrates the preciseness of
Bantu speech.
(ii) The ki and vi Particles of Manner. — In Swahili adverbial
function may be imparted to a noun or a neutral root by these particles,
and occasionally by u. (Cf. Nyanja chi and zi, and Ganda bu.) e.g. : —
Kusema tn'zuri. To speak well.
Kuamkia /rfzungu. To greet in English fashion.
Enda wpesi. Go quickly.
(6) Definiteness to Nouns.
(c) Emphasis on one of two Adjuncts.
(d) Nominal or Pronominal Function to Neutral Roots.
These may all be implied by the use of the Initial Vowel in LuGanda,
a full discussion of which will appear in a later issue of the Bulletin.
For the present it must suffice to say that the implication of definite-
ness plays a large part in Bantu speech. By this is meant more than a
mere the " idea of definiteness. Words such as ‘‘ my” or ■' this .
used as an adjunct to a noun, give to that noun an implication of
definiteness, and in LuGanda the noun assumes the Initial Vowel,
e.g. Pkitabo kyange (My book). Should the adjunct of the noun
imply indefiniteness, the Initial Vowel is not required, e.g. Nitabo ki ?
(Which book ? ). It is intere.sting to note in passing that the use of the
Objective Prefix with its Noun is but another example of this
implication of definiteness. Its inclusion or non-inclusion in a sentence
often puzzles a European, until it is realized that the implication of
the English wording supplies the key to its correct translation.
IV . Relationship Particles, a o e.
Before discussing these particles it should be clearly understood
that they are treated as affixes from a functional point of view.
STRUCTURE OF BANTU LANGUAGES
1117
Whether they are also “ roots ” is irrelevant, for the article deals with
function rather than with nomenclature. Nomenclature is, however,
of vital importance, and is beginning to receive the attention it
deserves. Our thanks are due to Professor Doke for his book,
published recently, dealing with Bantu terminology. ^ According to
the author, a is a formative ; he does not give it the status of a word,
and therefore does not allow it to be reckoned as one of the Parts of
Speech. Be that as it may, its function is prepositional, and with the
prefix proper to its particular context it translates almost every
preposition in English. Unfortunately o and e have escaped mention
by Professor Doke. The frequency wdth which a and o occur in almost
all Bantu languages (together with e in LuGanda) denotes how
fundamental they are, and stresses the necessity of getting a clear
understanding of their function.
(a) The “ a ” of Relationship.
The form which this relationship takes varies according to (a) the
nature of the words to which it relates, (^) the prefix which it assumes.
(1) a between a noun, and some other word or words takes the
concordial prefix of that noun, and helps to express an adjectival (or
“ qualificative ”) phrase.
Kiti cha mti. A wooden chair.
Kisu cha kukatia nyama. A knife for cutting meat. (A carving-
knife.)
Alimpiga kofi la chavu. He slapped his face. (He hit him a
cheek slap.)
(2) a between a verb and some other word or words prefixed by
ku (ku -f a = kwa) helps to express an adverbial phrase.
Alikw'enda kwa miguu. He went on foot.
Alishindwa kwa hofu. He fainted through fear.
Alistaajabu kwa ukubwa wake. He was astonished at its size.
(3) a between a noun and some other word wUen prefixed by n-
indicates an associative idea. In English this associative idea may need
to be translated by a conjunction or a preposition according to context.
Enda na Hamisi. Go with Hamisi.
IVIimi nawe. You and I.
(b) The “ o” of Relationship.
The relationship function of o (and of e) is that of pronominal
reference to some other wmrd in the sentence, expressed or understood.
^ Op. cit.
1118
E. O. ASHTON
It frequently requires no equivalent in the English translation. Hizi
= these + o of reference = Hizo. The use of o is the embodiment of
the Bantu characteristic of preciseness. Whenever there is reference
to something already mentioned or implied, in Bantu generally and in
Swahili particularly, o appears as a suffix (and occasionally as a prefix
also, e.g. ovyo = just anyhow). In Swahili one of its many uses is as
a particle of reference to indicate the relationship between a Noun and
its Antecedent in a relative sentence. For further details of the o
particle in Swahih, see “ The ‘ Idea ’ Approach ” already mentioned.
(c) The “ e” of Relationship.
This e prefix is characteristic of LuGanda. Its function is twofold :
(i) it corresponds to the o of Swahili ^ as the particle of relationship
between the Object and its antecedent in a relative clause, e.g. Ekitabo
kyenjagala (The book I want) (Sw. : Kitabu nitakacho). (Note that
the Initial Vowel in LuGanda acts as the relationship particle between
the Subject and its antecedent.)
(ii) e in LuGanda also relates the subject of a sentence with its
complement, if that complement is nominal. It is known here as a
‘‘ Copula ” in Bantu terminology, e.g. Kye kibira (It is a forest).
V. Formative Affixes.
(a) In Relation to Noun Forms.
These suffixes, which indicate a change in the form of a noun,
do not vary very greatly. Some languages are richer than others.
Swahili, for instance, has six noun suffixes, a, e, i,ji, o, u. Space forbids
more than a cursory mention of them. These suffixes are dealt with
in some grammars and ignored in others. They may be briefly described
as follows : —
(i) -a with the Living Class Prefix indicates a personal agent. It
is generally followed by a noun used qualificatively, e.g. Mshona
viatu, a shoemaker.
-a with the hi- prefix generally implies impersonal agency, e.g-
Kichinja mimba, a slayer of pregnancy (i.e. the youngest born)
Kinyosha mgongo, a back straightener (i.e. a tip).
(ii) -i also denotes an agent, without qualification, e.g. Mshoni,
one who sews.
It is probably correct to surmise that e in Swahili has disappeared, surviving in
singular of the Living Class only, and that o has taken its place.
STEUCTUEE OF BANTU LANGUAGES
1119
(iii) -ji denotes an habitual agent,' e.g. Mwindaji, a hunter.
Msomaji, a reader.
(iv) -e indicates resultant state. It is probably connected with the
e of state in verb forms in languages other than in Swahili, e.g. Mkate,
something cut off (hence a round of bread, a plug of tobacco). Kiumbe,
something which has been created. Mtuine, one who is sent (hence an
apostle).
A few of the so-called adjectival root.s take this suffix -e. e.g. eupe,
white ; nenc, stout. Thus used, they describe state.
(v) -0 with prefix m or nm indicates : —
Action. Mchezo, a game.
Result of action. Mwanzo, a beginning. 3 /apatano, agreement.
d/apatanisho, reconciliation.
Place. Malisho, a feeding ground.
-0 with prefix ki indicates instrunmit.
A'tzibo, a stopper.
(vi) -u indicates state or quality, e.g. Upofu, blindness. Kipofu,
a blind man. Hence u appears in many of the so-called adjective roots
which describe state, e.g. -ivw, ripe ; -ov«, rotten ; ref«, long.
These suffixes are dealt with very fully in Dr. Alice Werner's
Introductory Sketch of the Bantu Languages.
(b) Derogatory Idea in Relation to Nouns.
Most Bantu languages employ an affix to give a derogatory idea
to a Noun or to imply an insult. In Swahili is prefixed (or ‘‘ infixed ”),
e.g. vi/ipesa (a few worthless pice). In Zulu, with nouns indicating
females, the suffix -kazi is employed, e.g. umfaziA'azi (a great hulk of
a woman). It should be noted, however, that in Swahili ji has other
uses which do not imply a derogatory idea.
(c) In Relation to Verb Forms.
By means of suffixes various phases of verbal activity are imparted
to a Verb. The resultant forms are commonly spoken of as the “ Verbal
Derivatives ". Bantu languages are usually very rich in these suffixes.
Swahili, however, has few beyond the most fundamental ones.
(i) Simple -a (v) Passive -wa
(ii) Prepositional -ia, -ea (vi) Causative -ya, -vya, -fya,
-lia, -lea -za, -sha
^ Formed probably from -i attached to the >f/« verbal suffix of habitual or con-
tinuous action (now obsolete in Swahili).
1120
STRUCTURE OF BANTU LANGUAGES
(iii) Neuter -ika, -eka, -uka (vii) Positional -ma
(iv) Associative -na (viii) Conversive -ua
In ChiNyanja there are affixes to indicate other aspects of the
Verb, such as : —
Completeness, Insistence
-ta
Interval of time
-ba
Obligation
-dzi
“ Just in time ”
-fa
Mereness
-ngo
In Zulu the suffix -isisa indicates intensiveness. In Swahili this
idea may possibly be seen in the words saa and sana. In Swahili,
further, an old adage runs : “ Fungaio haiumizi mkono " " Binding
well does not hurt the hand ”, which is ob\dously an echo of the ta
of completeness. And again one may easily connect such words as
futnhata (to grasp) ^ with this same particle.
The function of many of these secondary verbal suffixes appears
to be adverbial, or, according to Doke, descriptive. They could equally
well appear under the heading of Implicatory Suffixes. A study of
these adverbial affixes in relation to verbs would bring to light many
interesting points. The above illustrations do not touch even the
fringe of such an investigation. It is interesting to note how
those languages, which in process of time have shed these adverbial
affixes, make good their loss. Thus in Swahih there is no adverbial
suffix to express “ to have just done something ”, but the same idea
is expressed in the makeshift phrase “ Ndio kwanza ” followed by the
verb in the subjunctive, e.g. Ndio kwanza wafike, they had just
arrived.
VI. CoNJUGATIONAL AFFIXES IN RELATION TO VERBS.
(а) Time particles.
(б) Aspect particles. These require a chapter to themselves and
will be discussed in a later article. They are dealt with in brief outline
in “ The ‘ Idea ’ Approach
(c) Mood suffixes. The Subjunctive Mood with its suffix -e must
also be left over for the present, together with the —
(d) Negative particles ha, i, si.
(e) The prefixes of person and number in relation to the verb call
for no particular comment.
* Fumba, to close ; fumbata, to take or grasp in the hand.
The e and o of LuGanda and the o of Swahili
By E. 0. Ashtox
T T would seem that originally there were two particles in Bantu
to indicate pronominal relationship of reference, but that in
process of time some languages (among them Swahili) dropped the
e and allowed o to usurp some of its functions. The e has been retained
in LuGanda and plays its own particular part, o, too, occurs in
LuGanda, partly coinciding with its use in Swahili. It also occurs in
LuGanda in places which in Swahili would require a different
construction.
e and o are alike in that their syntactical function is the same —
that of reference to some word expressed or understood. Each requires
a concordial prefix, and each can be attached to various parts of speech.
1. The 0 IN Swahili
It appears as : —
1. A relationship particle between the Subject and its antecedent,
e.g. Kiti kilichoanguka, the chair which fell down.
2. A relationship particle between the Object and its antecedent,
e.g. Kitabu nitakacho, the book which I want.
3. A relationship particle between the Demonstrative expressed
and its antecedent imderstood, e.g. zizo hizo, those very same ones.
4. The adverbial complement of manner of “ ndi ”, e.g. NdiCT/o,
that (in mind) is how it is done.
5. The pronominal complement of ” ndi- ” in relation to some-
thing in mind, e.g. Ndicho, that's the one.
6. Adverbial suffix of Place to the verb to be e.g. AlikuwaA'o,
he was there (i.e. place in mind).
7. Pronominal complement of the na of association, e.g. Ninacho,
I have it (i.e. something in mind).
8. Pronominal complement of the kica of instrumentality, e.g.
KwacAo, by means of it.
II. The o in LuGanda
In LuGanda o performs the work of 3, 6, and 7, leaving that of
1 to the Initial Vowel, and that of 2 to e. In addition o in LuGanda
is found as ; —
1. Suffix to the Partitive ku, e.g. FumbaAo, cook a little of it.
1122
E. O. ASHTON —
2. Expressing relationship of inanimate things, e.g. Ekai kyaA'^o,
the fibre of it.
3. Adverbial suffix to finite verbs in reference to place understood,
e.g. SaliraM'o, cut it there. Gyawo ebintu, take the things away
(from here).
4. Complement to the ti of likeness, e.g. Jiwetyo, like that (in
mind).
5. Particle of relationship in such phrases as ; Ekintu kinabyo
ebintu, one of them. Binabyo, the others. Kinabyo, its fellow.
6. Noun formativ'e meaning place to do something in, e.g.
E’somero, a place to read in. Ediro, a place where food is eaten (cf.
malisho in Swahili).
III. The e in LuGanda
e, like o, is a particle of reference, but its use is restricted : —
1. It indicates relationship between the Object and its antecedent,
e.g. Omuntu gwcnjagala, the man whom I want. It should be noted
that in adverbial expressions the antecedent is more often implied
than expressed, and its concordial prefix supplies the key to the
implication. Also the e often coale.sces with the following sound under
certain conditions, e.g. Lwalija ( < lu + e + alija) ombulire, when
he comes tell me. The concordial prefix lu agrees with olunaku = a
day. Wano wentambula ( < wa + e + ntambula), here where I am
walking,
2. It also indicates relationship between the Subject and a nominal
complement. In Bantu phraseology it is a “ copula ". e.g. Ye mulenzi.
it is a boy. Kye kibira, it is a forest. In LuGanda a relative construction
is sometimes used to give emphasis. This explains such sentences as ;
Wetuli, we are here (lit. : it is here we are). A sentence such as " I
want some boohs (not paper) ” would require the use of the e : Ebitabo
byenjagala, it is books which I want.
These foregoing examples of the o and e, when put into tabular
form, show clearly that their syntactical function of reference relation-
ship is one and the same : —
Swahili.
Hicho
Nacho
Yupo
Nilichotaka
Niendako
LuGanda.
Ekjm
Nakyo
Aliwo
Kyenjagala
Wentambula
THE E AND 0 OF LCGANDA AND THE O OF SWAHILI 1123
Funibako
Gy a wo
B wet VO
Kinabyo
Kinakyo
Ekai kyakyo
Kilichopotea
Ndicho
Xdivyo
VOL. Vm. TAUT t.
REVIEWS OF BOOKS
Sinica
By E. Edwards
My Coustey and My People. By Lin Yutang. pp. xviii + 363.
London : William Heinemann, Ltd., 1936. 15s.
Dr. Lin is that too rare phenomenon, a patriot with a sense of
humour. And because he has humour he has courage, and does not
attempt to conceal the weak places in the armour of his country.
Indeed he seems at times to flaunt them a little, as if to show how
confident he is that they will not in the end be able to bring about
her defeat. And yet he is not really boastful, nor lacking in clear-
sighted appreciation of the magnitude of China's problems and of the
fact that she alone can set her house in order. No more human book
on China has yet, to my knowledge, appeared.
“ How is China to be understood ? IVho will be her interpreters ? ”
are questions Dr. Lin puts to himself and his readers. And he follows
these with the still more pertinent inquiry, Do the Chinese under-
stand themselves ? Will they be China’s best interpreters ? ” How
few of our friends see our families as we do, or are able to accept our
estimate of them ! But sometimes one more understanding than
the rest is able to bridge the gulf between his home and the rest of
the world, and to suggest, by his frankness, his consideration, and his
sincerity, that his family must be “ nice China has such an interpreter
in Dr. Lin. It does not matter whether we accept at its face value
everything he says of China, or whether we agree with his theories
and behefs about his country’s past, present, and future ; here is
a Chinese, a member of the great Chinese family, who can make us
understand why exercise is anathema to them, ivhy funerals are
occasions of merrymaking, and why they have indomitable patience,
a less sensitive nerve-system, and yet, apparently, less courage than
peoples of other “ families ”. “ I write only for the men of simple
common sense. ... To these people, who have not lost their sense
of ultimate human values, and to them alone I speak," Dr. Lin declares.
Such people are not few, even in these days, and Dr. Lin's honesty
has gained for his country more understanding sympathy than any
amount of “ whitewashing ’’ could have done. No problems are
solved by this book, no constructive theories set forth ; on the whole
its tone is pessimistic, but it is the truth as seen by one who “ has
1126
REVIEWS OF BOOKS
not given up hope and is still confident that China will, as she
always did, right herself again In leaving the reader sharing its
author’s hope and confidence the book will serve his purpose.
China. A short Cultural History. By C. P. FiTZGER.iLO. pp. xx ~
615. Plates and Map. Edited by Profe.ssor C. G. SELiGiUN, F.R.S.
London : Cresset Press, 1935. 30s.
This new history of China is intended for the general reader.
It covers the whole course of her development from prehistoric times
to the latter part of the nineteenth century. Its seven parts deal
with the seven great epochs of Chinese history, reducing to a minimum
the complications of political and dynastic events and treating more
fully cultural and economic dev'elopments. A .special feature of the
work, absent from many earlier histories of China, is the proportion
of space given to the period before the beginning of her relations with
the West in the nineteenth century. Art, religion, economic and
political experiments, and literature are discussed in connection with
each epoch ; there are many illustrations, carefully chosen to illuminate
the text, and Mr. Fitzgerald has made excellent use of the material
provided by modern scholarship and research, both Chinese and
Western.
It is not to be expected, in a work of this size, that all the parts
will be equally good. But it would be unfair to point to this or that
weakness without at the same time pointing out special excellences.
Speaking generally, I find Mr. Fitzgerald most interesting on questions
relating to social and economic conditions ; others of his readers
will, I imagine, be particularly interested in his discussion of the
drama and the novel, which have previously been but scantily treated.
In any case this is a book which may be warmly recommended to
the ordinary reader, for whom it is intended.
The New Culture in China. By Lancelot Forster. London :
Allen and Unwin, 1936. 7s. 6d.
In a somewhat pessimistic series of essays Professor Forster,
who, as professor of education in the University of Hong Kong, may
be said to occupy a position half-way between east and west, asks
many questions to which no immediate answer can be found. “ Is
SIXICA
1127
the old classical system dead beyond recovery 1 " Can Confucianism
be revived . . . and made to operate effeetively amon^ all classes of
the community On the whole he himself seems doubtful of
the effectiveness of such a revival, while sure that present experiments
in education without an ethical background are doomed to failure.
He describes China as “ without direct guidance and without principles
to which it can whole-heartedly adhere ", and says that until she
finds a solution of her philosophical problems '' efhciencv’ and discipline,
strength and security, wealth and national power will not be achieved
How her problems are to be solved neither he nor anyone can .sav
as yet. In addition to the new culture manv related topics are discussed
the mass educational movement, a social and educational experiment
of great importance to China ; the " Three Principles " of Sun Yat-
sen ; and other subjects which from time to time have aroused his
interest.
The varying moods in which the author writes reflect the successe.s
and failures of new ideals and revolutionary developments. China
has never thought in terms of years or decades ; if she seems slow to
grasp the consequences of applying the externals of western ci\-ihza-
tion to a great country whose habits of thought remain largelv
unchanged, the impossibility of waiting till the change has been
brought about from within before applying them is obvious to her
leaders, and the " bewildering sense of ferment " left, as Sir Michael
Sadler observes in his Introduction, in the mind of the reader, is a
true reflection of the struggle to make use of both methods in bringing
China into line with the rest of the world. In Professor Forster’s
view one of her fundamental needs is the destruction of the idea of
the importance of the individual, and the merging of the welfare of
the individual in that of the wider political group of which he is a
member. In view of her social traditions, in which the familv is the
unit, the transference of the emphasis from the family to the State is
a problem which goes to the root of the effective political organization
of the country. The enormous size of China makes the task one of the
greatest difficulty; when she can teach her people that " unitv i?-
strength " the purpose of the revolution of 1911 will have been
achieved.
A wide circle of readers to whom the situation in China is as
puzzling as it is kaleidoscopic will welcome Professor Forster’s effort
to elucidate for them, in a sincere and straightforward manner, some
the more intricate bits of the puzzle.
1128
REVIEWS OF BOOKS
A Pageant of Asia. A Study of Three Civilizations. By Kenneth
Saunders, pp. 464. Oxford University Press, 1934. 21s.
The Ideals op East and West. By ICenneth Saunders, pp. xxiii,
268. Cambridge University Press, 1934. 10s. &d.
In both these volumes Dr. Saunders elects to cover a wide field,
for, like other writers of the present day, he believes, after personal
contact with contemporary scholars and thinkers of India, China,
and Japan, that the civilizations of Asia should be studied as a whole
rather than as a number of single units. One of the drawbacks of this
approach is that it involves so many omissions. Dr. Saunders' pageant
includes only certain of the great ages of each of the three countries
of which he treats, and emphasizes particularly the religious and
philosophic aspects of their development. He lightens his task and
assists his readers to become familiar with the thought of the sages of
old by quoting many passages from the literature of each country.
This method he follows to an even greater extent in The Ideals of
East and West, whose aim is “ to be useful in an age of transition,
when ethical ideals, like everything else, are being tried and tested
Originally delivered as Earl Lectures at the Pacific School of Rehgions
in Berkeley, California, these chapters on the ethics of the chief religious
systems of Europe and Asia depend to a considerable extent upon
the illustrative readings which make up something like two-fifths of
the whole.
“ Not only in the parallels and similarities between these great
systems but in the contrasts which emerge from a comparative study
is there useful matter for thought and conduct,” says the preface.
There is perhaps at times in Dr. Saunders’ attitude a tendency to
create in the minds of those of his readers who have not studied the
subject an impression that the similarities between Taoism and
Christian ethics for instance are closer than in fact they are ; the
Christian ideal of the Holy Spirit working in the heart of man to
guide his footsteps in the Way of God and the spontaneous existence
depicted in the Too te citing have little in common.
Dr. Saunders is to be congratulated on the selection of his
anthology, and both books may be recommended to the general reader
provided he is not misled into supposing that the similarities which
Dr. Saimders points out go farther and deeper than Dr. Saunders
himself means to suggest.
SIXICA
1129
Men and Gods in Mongolia (Zaa'agan). By Henning Haslund.
Translated from the Swedish by Elizabeth Sprigge and Claude
Napier, pp. xvi + 358. London : Kegan Paul, 1935. 15s.
For more than three years (1927-1930) Mr. Haslund was on
active service with Sven Hedin's Central Asiatic Expedition. The
main features of that journey have been published already, but the
personal experiences of the members of the expedition have their
own special interest.
Mr. Haslund's first book. Tents in Mongolia told the story of
his introduction to Mongolia. In his own straightforward style he
related how he and others were foiled by pohtical disturbances in
their efforts to found a farming colony in a Httle-known district of
Mongolia. The feature of the book was the author's own power of
finding the common denominator of humanity, and of sharing with
youthful enthusiasm and ready sympathy in the lives of the strange
people whom he met.
The present volume finds him no less enthusiastic and sympathetic,
though more mature, and the events narrated are not less exciting
than his earlier experiences. Across the desert to Estiu-gol, through
the middle of the terrible Black Gobi, to Hami, where his whole
party was arrested, on to Urumchi under armed guard, he came at
length to the country of the Western Torguts, whom he had in the
beginning set out to find. The most important part of the book is
the account of this mysterious people, among whom the author lived
for a long time, investigating their customs and their laws and
exploring the surrounding countrj'.
As a Shaman and the friend of a reincarnation of the Buddha he
was able to learn much that is of value to the ethnologist ; as a
traveller he will be read with delight by all who enjoy romantic
adventure.
The book is well illustrated with photographs and the words and
music of Torgut songs and choruses. There is also a map showing
the author's journejdngs from 1927 to 1930.
Confucianism and Modern China. By Reginald F. Johnston,
K.C.M.G. pp. 272. London : GoUancz, 1934. 8s. 6rf.
Sir Reginald Johnston's book is based on lectures delivered at
Bristol University in 1933. The signs w'hich he then perceived of
the rehabilitation of Confucianism have since been greatly magnified.
1130
REVIEWS OF BOOKS
China’s leaders have realized that a cultural background cannot lie
changed at will like a theatrical back-cloth, and that a new culture
niu&t grow, just a.s the old one grew, out of emuronment and character.
To many serious Chinese the dangers of abandoning Confucianism
have been evident from the beginning. W'e may well doubt, as Sir
Reginald -John.ston does, the sinceritv of Chinese who profe.ss to think
We.stern civilization .superior to their own. Discarding the .system
while holding to the tradition, and adopting a new and. in their
opinion, inferior culture could have onlv one result. The decksion
to ob.serve the birthday of Confucius as a national holiday inean.s
that what may have .seemed like undue optimism on the part ot
.Sir Reginald in 1934 has been entirelv justified by the course of events
during the la.st two years. iModification.s of Confucianism there cannot
but be ; but the innate power of resistance to outside influences
which has In'cn a markerl characterisfic of the Chinese people through-
out their hi.story must in the long run prevent them from denying
the truth that is in them ; and the e.ssentials of her old tradition
must and will form the foundation of the structure of the new China.
A.s Lin Yiitang .says, ".She will, as she always did, right herself
again.”
The Wav a.vu us Rower. A study of the Tao Te Ching and its
Place in C’liine.sc Thought. By Arthtr W.vlev. pp- db-.
London; Allen and Cnwin, Ltd., 1934. 7s. 6f/.
Mr. Walev fives tlic date of the Tno fe chine/ a.s the third century
n.c. atid interprets it with that period as its background. The result
is revolutionary ami illuminating. .\sa work of the si.vth century n.c..
in which its supposed author, ijao Tzu, is said to have' lived, much of
the Tno tc cJiim/ was incomprehcnsilile : set in relation to all the
other schools of thought which e.visted in the third century it is seen
to be a collection of .saying.s l)orrowed from other and often opposiiig
.systems of belief and ingeniou.sly turned to suit the author's purposes.
Mr. W ah;y s translation aims ” to reproduce what the original .say.'
with detailed accuracy ”, and he has added paraphrases and com-
mentaries further to elucidate the meaning of tliflicult chapters.
But tins translation and tluvse commentaiies would still leave much
of the Tao te chin;/ unintelligible without the Introduction, for which
Mr. \N aley is inclined to apologize because it comprises more than
half the book. But it is for this introductory studv, intended for
SIXICA
1131
those who liave no professional interest in Chinese studies, and the
appendixes, intended chiefly for specialists, quite as much as for his
historical translation that we are indebted to Jlr. Waley, who well
knows how to set scholarship at the service of the ordinary reader.
The Literary IxQuisinox of Ch'ien-Llng. By Luther Carri.\(;ti).\
Goodrich, pp. xii -J- -l-l- Baltimore : Waverlev Press, 1935. 15'.
This volume is Number 1 in a series of studies in Chinese and
related civilizations, published for the American Council of Learned
>'Ocieties. Mr. Goodrich, who is lecturer in Chine.se at Columbia
Lniversity. has undertaken in it a study of the literary censorship of
the eighteenth century in China. Such censorship was not uncommon :
despotic emperors from t'ne time of Ch’in Shih Huang-ti had known
how to force their will upon contemporary scholarship. Mr. Goodricli
uives a brief resume of book-burnings prior to 1772. the year in
which Ch'ien Lung issued the eebet which brought pouring into Peking
l>ooks of every sort, ” meritorious " and otherwise. Among them were
a considerable number deemed to be subversive and defamatory to
the Manchu dynasty. The result was the compiling of two biblio-
graphies — the S.^u k'u ch'>ia» •'/ot Catalogue, containing all books
worthy of a place in the imperial library, and an hulex Expioyatorix^:.
on which were listed the books banned wholly or in part. Books were
suppressed for various reasons, of which Mr. Goodrich cites eight
principal ones. A large number, written at the end of the Ming
dynastv. were anti-iManchu ; others defamed earlier dynasties or
peoples claimed by the Manchii.s as ancestral ; geographical works
.sometimes contained information conshlered dangerous to the ruling
house, ami so on. The net was large and its meshes very tine.
After discussing the Inquisition in part i, iMr. Goodrich proceeds,
in the second and longer part of the work, to translations of the
biographies, memorials to the throne, imperial edicts, and proclama-
tions on which his study is partly based.
It is odd that the true .signiticance of the Indv.r. whose existence
has long been known, was not discovered earlier, and our thanks are
due to iMr. Goodrich for his careful and well-annotated study of an
extremelv interesting subject.
1132
REVIEWS OF BOOKS
The Romance of the Western Chamber (Hsi hsiang chi). A
Chinese play, translated by S. I. Hsiung. Preface by Gordon
Bottomley, pp. xxiii + 281. London ; Methuen, 1935. 8s. 6d.
Chinese stage conventions emphasize what Chinese ethics condemn.
Plays about clandestine love alfairs are no doubt a natmal result
of the strict segregation of the sexes enforced by Confucian practice.
Such plays, like “ gangster ” films, are meant to be enjoyed biit not
imitated. It would not be just to blame the “ barbarian ” Mongols,
under whose rule the Chinese drama blossomed so richly, for this
characteristic of the theatre. The Western Chamber and other plays
with similar themes are based on stories written during the T’ang
period (a.d. 618-906), under a Chinese, and not a foreign, dynasty.
Mr. Hsiung has translated this “ really artistic ” drama to prove
to us that Lady Precious Stream was “ just a popular, commercial
play ”. Some comparison of the two plays would thus seem to be
invited in order to discover precisely in what the superiority of the
Western Chamber consists. Their conventions (which may account
for a good deal of the popularity of Lady Precious Stream) are equally
odd, and there is little to chose between them in unreality. If the
“ artistic ” play were reduced by the omission of the songs we should
probably find in what remained that both in theme and characteriza-
tion the “ commercial ” play could hold its own ; and it must be
confessed that the lofty sentiments so often expressed in the Western
Chamber are put into practice in its rival. In two particulars, none
the less, the Western Chamber is supreme : its elegant language and
its delicate imagery prove it the child of a scholarly mind, a Chinese
scholarly mind, which used the story as material on which to embroider
fair and fanciful designs in words. This being so, it is even a little
shocking to the reader to find that the translator has sometimes
chosen to give the English a “ quaint ’’ or a comic turn, admirably
suited to Zody Precious Stream, but not to the more “ classical ”
style of the Western Chamber. Nevertheless iVIr. Hsiimg’s translation
deserves our thanks and will be widely read.
Annual Customs and Festivals in Peking, as recorded in the
Yen-ching Sui-shih-chi. By Tun Li-Ch‘en. Translated and
annotated by Derk Bodde. Peip’ing, 1936.
This is not the first book to deal with the customs and festivals of
Peking ; but it is the first translation of the annual cycle of fife in
SIXICA
1133
Peking seen through the eyes of a Chinese. The author of the original
work. Tun Li-ch'en, was a Manchu, born in 1855, who spent most
of his life in Peking, which he found so full of interest that he amused
himself by writing this record of customs, festivals, and ceremonies
and published it in 1900.
Mr. Bodde deserves our gratitude for introducing us to this old
Manchu gentleman, and through him, to the everyday life of the people
of Peking and of China, for many of these customs and feasts are not
peculiar to any one place.
Students of folk-lore and sociology as well as students interested
specially in China will find in this volume much that wdl appeal to
them. There is, perhaps inevitably, a certain monotony about
descriptions of feasts and ceremonies, but the additions of the compiler,
sometimes scholarly, sometimes just “ auld wives' tales”, help to
maintain the reader’s interest.
Mr Bodde has added useful appendices relating to the calendar
and other matters. Chinese characters are used throughout the book
both in footnotes and for the names of festivals. Some of the illustra-
tions, which include sis plates, twenty -eight text-drawings and a map
of Peking, are reproductions of old wood eng^a^■ings, and the remainder
are the work of a modern Chinese artist.
To those readers who knew China before the days of industrial and
political revolution this book will remve old and pleasant memories ;
those who know her in her present days of stress and hiirry will find
in it some explanation of the essential simplicity, due to close
contact with nature, benign, or starkly cruel, of the Chinese people as
a whole.
Chinese Festive Board. By Corri.nne Lamb, with line-drawings
and paper-ends by John Kirk Sewall. Peiping ; Henri Vetch,
1935.
As befits a properly civilized people, the Chinese have always
regarding eating as an art as well as a necessity. In spite of his
declaration that the ” earnest scholar does not demand his fill when
he eats ”, Confucius invariably refused to eat meat that was not
minced to his liking, required ginger with every meal, would not sit
on his mat if it were not straight, and did not stint the wine. In the
days when Alfred was burning cakes in England, Chinese scholars
were in the habit of giving elaborate dinners when they successfully
1134
REVIEWS OF BOOKS
passed their state examinations. The menu of one of these feasts,
which has come down to us from that period, includes many dishes
with fanciful and fantastic names. ■’ Longe\-ity gruel explains
itself; ■■ white dragon-brain " was camphor, which the Cambodians
also ate at feasts ; " fairy-meat ” consisted of slices of chicken cooked
in milk. Other dishes popular at that time and obtainable in special
food-stores were " unborn phoenix ”, " ivory dumplings ”. and " broiled
dragon's whiskers ilrs. Lamb sets none of these delectable (and
probably unpalatable) dishes on her Chinese Festive Board, but she
offers to EngHsh-speakin" enthusiasts for Chinese food — and they
are many — a varied and attractive selection of recipes which can be
prepared in foreign as well as in Chinese kitchens. Of more general
interest are the introductory pages in which table eticpiette, liquors,
and drinking-games are discussed. An acquaintance with these and
similar conventions should be helpful to the uninitiated Westerner in
China.
Chinese Art. Edited by Leioh .\shto.v. pp. xvi - 111. London:
Kegan Paul, 193.5. 5s.
Chinese Art. An Introductory Handbook, pp. 102. London :
B. T. Eatsford. Ltd., 1935. 15.>.
A Backgrol'Nd to Chinese Painting. By .Soame .Jenyns. Assistant
Keeper. Department of Oriental Antiquities, British Museum,
pp. 237. London : iSidgwick and .Jackson, Ltil., 1935. lO-y. Ctd.
The Chinese Eye. An Interpretation of Chinese Painting. By
Chi.\ng Yee. 24 plates, pp. .xvi 240. London : 3Iethuen.
1935. 7s. 5d.
Introduction to Chinese Art. By Arnold Silcock. pp. 285.
London ; Humphrey Milford, 1935. 6.y.
The International E.xliibition of Cliine.se Art at Burlingou Hou.se
in 19.35-6 was responsible for the publication of a considerable number
of books dealing with one or more aspects of Chinese art. It is not
an easy matter for the reviewer to determine the particular usefulness
of each. The majority were intended, obviously, for readers interested
rather than informed in Chinese art. anil their authors have there-
fore devoted considerable space to background and fundamental
principles. Among works treating of Chinc.se art, a .small boo.k edited
by Leigh Ashton with articles on painting and calligraphv, .sculpture
SIXICA
1135
and lacquer, the potter's art, bronzes, jades, and textiles, mav be
given a foremost place. Mr. Binyon contributes an introduction and
discusses tbe twin arts of painting and calligraphy. Xo one knows
better than Mr. Binyon how. by an image or a phrase — cf. p. 2. “ you
will find yourself taking the same sort of pleasure ' ' (in the l)rushwork
of calligraphy) that one has in the clean sharp shapes of iris-blades
thrusting up from the sod, or the swift swerve of water past a stone "
— to give direction to imagination and .stimulate the mind to transfer
the emotional experience of IVe.stern art to the investigation of Chinese
art ideals. In a chapter entitled '■ .Sculpture and Lacquer the
Editor condenses into thirteen pages the facts essential to an examina-
tion of actual pieces, and in an even fewer number of pages he deals
briefly but adequately with textiles. Similarly Mr. K. L. Hobson
takes his readers over the whole ground of the potter’s art, using
broad strokes to paint his picture, but filling in details wherever they
are hkely to help and not confuse. Mr. A. J. Koop and Dame Una
Pope-Hennessey, both specialists in their respective subjects, are
responsible for articles on Bronzes and Jades. The book is pleasingly
produced, suitably illustrated and very chea]). and for the not too
serious amateur must have served as an admirable guide, while, too,
it should still be an excellent introduction for those who wish to
know more about the subjects of which it treats.
Unlike the essays which compose Chinese Art. Mr. Silcock's book
does not assume any previous knowledge of tlie subject. To make
things easy for beginners he avoids names and dates as far as possible,
atoning for thi.s omission by adding as an appendix a series of .synoptic
tables giving the dates of important persons and event.s in Chinese
history with concurrent events in the rest of the world. Most of the
material of this effective introduction has been gathered from the
works of experts in each field, to whom Jlr. Silcock acknowledges
his indebtedness. The lucid prc.sentation of the material thus obtained
is his own, and he succeeds admirably in providing ” a simple picture
of the environment and the age-long ileveiopment of a great people
and a noble art ",
The illustrations, which comprise sixteen text-figures and twenty-
seven plates are admirably chosen and the volume is very pleasingly
produced.
In a quite different category is Mr. .lenyn’s Bnclyround to Chinese
Paintinff. The introduction, by Mr. \\ . AV. M’inkworth. is addressed
“ To Collectors ". and one feels that the author had in mind an audience
1136
REVIEWS OF BOOKS
at least vaguely familiar, aud not that unknown quantity, the “ general
reader After a general survey, for the purpose of which he divides
painting into seven periods, Mr. Jenyns deals with the influence of
religion on painting ; its relation to calligraphy ; materials and
technique ; the treatment of landscape and the human figure ; and
the use of bird, flower, and animal motives. Opinions will differ as
to which of these chapters is best ; this is not a book for an unsophis-
ticated public, and those to whom it is addressed will decide the
point for themselves according to the special aspect of Chinese painting
which appeals to the taste of each. The numerous illustrations, some
of them unfamiliar, are a delight, and the book is extremely well
produced.
In The Chinese Eye, IMr. Chiang Yee, himself an artist, writes of
Chinese painting from the Chinese point of view. He discusses not
only the history, the essentials, the instruments, and the species of
painting, but also the relation between it and philosophy and literature.
Mr. Chiang’s own paintings, in a variety of styles, have been twice
exhibited in London, and those who saw them will be especially
interested to learn, from his own pen, what are the emotions and the
principles which imderly his work and that of his fellow-artists past
and present.
On the technical side Mr. Chiang is instructive ; his translation
of Hsieh Ho's Six Canons of Painting is interesting and worthy of
note ; on the artistic side his is the true poet’s approach, and he
illuminates for us every now and then the attitude and the artistic
conceptions of the Chinese painter. Speaking of the inscriptions
found on most Chinese paintings, he says, “ Veri-similitude is never a
first object ; it is not the bamboo in the wind that we are representing
but all the thought and emotion in the painter’s mind at a given
instant when he looked upon a bamboo spray and suddenly harnessed
his life to it for a moment. That moment may have had a psychological
significance in itself — the artist was . . . parting from friends — then,
the drooping attitude of the leaves wall have a meaning for their
creator beyond the comprehension of the onlooker had he no written
explanation.”
Mr. Chiang acknowledges his indebtedness to Miss Innes Jackson’s
generous help in rendering his work into a lucid English style and
drawing his attention to parallels in European thought.
The twenty-four illustrations are grouped under three heads —
Figures, Landscapes, and Birds, Flowers and Animals. They are
SINICA
1137
specially selected to illustrate various points tlrroughout the book
and are representative of the best of Chinese painting.
On a larger scale than any of these is Messrs. Batsford's Chinese
Art, published on the occasion of the Exhibition. This work is based
upon the Burlington Magazine Monograph, Chinese Art, published in
1925, long out of print. WTiile, unfortunately, some of the contributors
to the former pubhcation found themselves unable to assist in
the preparation of the new edition, the publishers are to be con-
gratulated upon producing a series of essays by well-known authorities
in the subjects treated. Mr. Lawrence Binyon writes on painting.
Dr. 0. Siren on sculpture, and Mr. Bernard Eackham on pottery
and porcelain. Mr. A. F. Kendrick and Mr. W. W. Winkworth also
contribute, while Mr. Eoger Fry writes on the significance of Chinese
art, and Mme Quo Tai-Chi has permitted the inclusion of a review
of forty centuries of Chinese art which she contributed to the Queen.
It is impossible, within the limits of a brief review, to discuss
these essays individually. It is sufficient to say that here is a work
which includes an outline of the historical background of the principal
forms of Chinese art, and an account of the important types in each
field. The book is intended to be “ popular ” ; it is at the same time
comprehensive and yet free from an over-elaboration of detail, easy
to read and yet informative. The pubhshers, feehng that “ however
illuminating the text, the most enlightening and satisfactory method
of appreciating Chinese art is by the study of comparative examples,
and especially by presenting these in the beauty of their original
colour ”, have generously packed the book with illustrations — 23 plates
in colour and 62 in monochrome, all beautifully reproduced.
The Chinese on the Art of Painting. Translations and Comments
by OsvALD Siren, pp. 261. Peiping ; Henri Vetch, 1936. 12s. 6d.
This volume is in some respects a complement to Dr. Siren’s
History of Early Chinese Painting, of which volumes i and ii, have
already appeared. The present work consists chiefly of translations
from Chinese works on painting, and was the result of a conviction on
the part of the author, which many will share, that a true knowledge
of the history and significance of Chinese painting must be based
on historical records and the works of Chinese critics rather than on
the comparatively few specimens which survive.
1138
REVIEWS OF BOOKS
In the arrangement of his material Dr. Siren has elected to follow
the sequence of the dynasties, believing that the process of evolutioi;
and the special ideals of each period can be thus most clearly presented.
As to the material selected for inclusion, here again Dr. Siren has
been content — and wisely so — to use what the Chinese themselves
have regarded as being of importance and have preserved in their
historical collections.
The greater part of the book consists of translations from sucJi
well-known works as Li tai »iing hua chi and Ku hna p'in lu. Sonu-
of the material had already been translated by others, but Dr. Siren
has co-ordinated the available information in relation to certain
problems, theoretical rather than practical, which he seeks to elucidate
as far as possible for the student. The book is illustrated by a number
of half-tones, well chosen and well reproduced, and includes an
adequate index. It will be well worth the while of the serious student,
whether of Chinese or of Chinese art, to make himself familiar with
the views of great Chinese critics and ma.sters of painting through
Dr. Siren's scholarly work.
Some Technical Terms of Chinese Painting. By Benjamin
March, pp. xiii -t 55. Baltimore ; Waverley Press, 1935. $1.50.
Increased interest in Chinese art and all subjects connected with
China makes more and more desirable the production of technical
works to aid western students. Many people whose enthusiasm tor
Chinese painting was fired by the recent International E.x:hibition m
London will welcome this handbook by a promising young American
sinologist, who died, unfortunately, before its publication. The volume
appears under the auspices of the American Council of Learned
iSocieties, as number two of the series entitled “ Studies in Chinese
and Related Civilizations".
The author's aim was to impart some knowledge of painting as
a craft, and his collection of terms is the outcome of a practical study
of the technique of Chinese painting made in Peiping under the
tutelage of a Chinese artist and student of the classical tradition.
The terms, which comprise the Chine.se characters, their romanized
equivalents, and an English explanation of their meaning, are grouped
for convenience under a number of heads, which include Materials.
iSubjects, Techniques, Bru.shwork and Composition, and Seals and
Signatures. The list is not exhaustive, even on the technical side ;
SIXICA
1139
still less does it include the terminology of aesthetics, but so far as
his study went, the author spared no pains to explain intelligently
rather than to define arbitrarily the meaning of the terms he collected
An alphabetical list makes the book easy to use, and a number of
plates illustrating brush-strokes and type-forms ” (i.e. the various
forms in which leaves, mountains, water, etc. are painted) add to its
interest and value.
Chinese Jade. By Frank Davis. Published privately by the
author, 1936. 5.s.
In this little book of less than seventy pages Mr. Da\-is has given
to the amateur of jade a reasonably accurate chart by which to
steer his course ”. He helps as much perhaps by what he omits as
by what he says, and what he says is the outcome of his own interest
in that friendly and noble substance *’ which intrigues him as it
intrigued the Chinese from the earliest times.
To write on jade after Dr. Laufer requires some courage, but
Mr. Davis has not tried to rival his predecessor but only to simphfy
the subject. He has, in the main, confined himself to an attempt to set
jade against its historical background, and to suggest its significance
in Chinese art and ceremonial. So little is known on this subject that
to a large extent theories must take the place of facts, and Mr. Davis s
theories are interesting, if not always convincing. Sometimes even
his facts are mis-stated, as for instance, when he avers that Chou
feudalism “ was at least a stable system, and it was enlightened
enough to give every honour to Confucius but we will not cavil
overmuch, for this sincere and concise little monograph deserves to
be welcomed by a large number of readers whose interest in the subject
is general and not speciahzed. The illustrations are well-chosen and
very well reproduced and the typography excellent.
Modern Newspaper Chinese. Progressive Readings with \ ocabu-
laries. Notes, and Translations. By J. J. Brandt, pp. xii + 321.
Peiping : Henri Vetch, 1935.
Many modern students of Chinese owe a good deal to Mr. Brandt s
earlier textbooks, Introduction to Literary Chinese and Chinese Particles ,
They will find additional reasons in this new work to thank him for
the painstaking manner in which he endeavours to solve the problems
VOL. Vm. PART 4.
1140
REVIEWS OF BOOKS
which beset the student who undertakes the study of newspaper style.
The stage at which students express a desire to be able to read Chinese
newspapers varies greatly ; if they begin their study of the subject
with this book they will cjuickly find themselves familiar, even in
their early days, with a wide vocabulary of terms and expressions
commonly used to-day. The author’s aim is to provide the reader with
interesting and up-to-date material for the study of the newspaper
style. His extracts, taken from the newspapers of Shanghai, Tientsin,
and Peiping during the year 1934, include every type of subject from
a Tax on Bobbed Hair to the Meaninr/ of Political Unification, and
reach from the children's weekly page to the resolutions of a financial
conference.
The sixty les.sons into which the books is divided are arranged in
progressive order of difficulty. Each lesson is accompanied by a
vocabulary and notes as well as an English translation. 31ost people
will agree that in order to translate into good idiomatic English it i.s
necessary to abandon in many cases the exact meaning of the Chinese.
Mr. Brandt, we are glad to observe, has chosen to stand by his student-s
and to offer them as close a rendering as possible of their text. For
this they will be especially grateful. It is not necessary that journalistic
Chinese should be rendered into literary English of a high quality ;
it is essential that the student should be able to see how the translator
arrives at his translation.
A useful handbook such as this will no doubt go to a second edition
when the author will be able to correct a number of typographical
errors. In the meantime students will be glad that they were no longer
deprived of the assistance which it cannot but give them in their
study of modern Chinese.
Chinese ^Made Ea.sy. By Walter Brooks Brouner and Fum!
Yuet iMow, M-ith an Introduction by Herbert A. Giles. Leiden :
E. J. Brill, 193.0. 10 guilders.
The present volume is no doubt a new edition, thoush there is
nothing save Professor Giles's Introduction, dated 1903, to indicate
that the book had been published earlier than the date printed on
the title page. And the date is important. In 1903 such a work must
have been of the greatest assistance to students, to whom few text-
ooks of the Chinese language were available ; in 1935 its usefulness
SIXTCA
lUl
is less, because in the interval the teacliiiig of Chinese in all its branches
has made great advances and good books and trained teachers are
not difficult to obtain. Nevertheless, for students of Cantonese the
book is not without value. It consists of a number of sections. Thirtv-
three lessons on various subjects form the first part. These comprise
sentences printed in character, with the Cantonese pronunciation
and the literal meaning of each character, as well as an English
rendering of the sentence. The second section includes the complete
text of the San tzu citing in large characters in the written style,
accompanied by a translation. Other sections include 1,000 characters
intended to give practice in writing and some sample pages of a Cliinese-
Enghsh dictionary. At the time of writing their i^reface the authors
claimed that their methods were unique but the student who buys
the book at the present time must not expect to find in it anything
new or revolutionary in the techniriue of language-study.
Chixese and Enollsii lIuDERN JIiLiTAKY DICTIONARY. 5,o00 army,
navy, air technical terms. Compiled by Captain J. B. Davidson-
Houston, R.E., and Lieutenant R. V. Dewar-Durie, A. & S.H.
(With illustrated appendixes showing naval and military badges
of rank.) Peiping : Henri Vetch, 1934.
The problem of translating western .scientific and other terms
into Chinese has been met by the establishment a number of years
ago of a special committee, with the result that a more standardized
vocabulary of suitable new words is being added to the Chinese
language than seemed possible before the committee came into being.
Early attempts to compile dictionaries of technical terms in Chinese
and European languages were hampered by the unsystematic methods
then practised by the Chinese themselves in translating western words,
and by the efforts of western specialists to express in Chinese technical
details of which the meaning was not understood by Chinese. Never-
theless even before the Great War many technical, medical, and
niilitary terms were included in Chinese-English and English-Chinese
general dictionaries, and a few special vocabularies of such terms were
compiled. Since that time, largely through the medium of the news-
papers, western ideas have become familiar to an ever increasing
number of Chinese, while civil wars and the modernization of the
Chinese army have helped to focus attention upon military matters
1142
REVIEWS OF BOOKS
in particular. The work which the compilers of the present handbook
have done, therefore, in preparing for the use of Chinese and Enghsh
speaking people a well-arranged and extensive vocabulary of mihtary
and technical language, cannot but be appreciated by all students
of modern Chinese.
Chinese Calligraphv. By Lucy Driscoll and Kenji Toda.
pp. ATu -f 71. University of Chicago Press, 1935. 9s.
It is related of Ou-yang Hsiin, the famous calligraphist of the
T'ang period, that he went to see an ancient tablet written by So
Ching of the Chin dynasty (a.d. 265-419). Reining in his horse, he
looked at it for a long time and then rode away. A hundred paces
olf he stopped, turned back, and remained gazing at the tablet until
he was exhausted. Then he sat on a blanket and continued to gaze.
He slept at the foot of it and remained there for three days.
Probably nothing has been, or can be written about Chinese
calligraphy which will explain fully this attitude of rapt devotion
before a specimen of fine handwriting. It is an interesting problem
on which a great deal more will be written.
The authors of the present work have gone direct to Chinese
sources for their material, and have endeavoured to interpret and
evaluate their quotations “ from a modern psychological point of
view ”. Though the method is attended by certain dangers, it is
interesting in itself. The literature of the subject is so large that it
has been “ barely tapped ”. The Shu fa (Laws of Writing) of Ou-yang
Hsiin (a.d. 557-641), which must surely have been the basis of the
“ Eighty-four Laws ” of Li Shun of the fifteenth century, is not
mentioned, though it is important both in itself and as one of the
earliest works on the subject, written at the beginning of a period
when caUigraphy was developed until more than fifty styles were
recognized.
Readers unfamihar with Chinese writing may find difficulty in
understanding some of the translations. The following passage and
the authors' comment thereon, taken at random (p. 51), is far from
clear ; —
When two characters are written so as to make one character,
the spots and lines, above and below, bending down and looking up,
should show the force of separating and uniting. . . . The order of
CAMBODIAX GLORY
1143
spots and lines ought not to be even and uniform ; it should show
unevenness like the arrangement of feathers or scales.’
“ He is speaking of the problem of breaking the integrity of two
movement patterns sufficiently to let energy go from one to the
other so that a unifying interchange is felt between the parts. . .
Neither will the student of Chinese, seeking a history of the art
of calhgraphy, find it here. The authors have made a praiseworthy
effort to look into, and interpret for those already acquainted with
the subject, the mind of the Chinese masters of calligraphy. The degree
of their success must not be minimized ; the book is interesting ;
we do see, from the interpretation offered, how the successful caUi-
graphist achieves a harmony between mood and material on the
aesthetic side, and how he attains a mastery over brush, ink, and paper
by a precision and a persistence which we usually associate with
scientific experiment.
The volume is well produced, and is illustrated by a number of
ink-rubbings from the collection of the late Dr. Laufer. It would be
much more easily handled, however, if an index had been provided.
E. Edwards.
CAMBODiAJf Glory. By H. W. Ponder, F.R.G.S., F.R.S.A. 84 x 5^.
pp. 320, plates 24, maps and plans 4. London ; Thornton
Butterworth, Ltd., 1936. 15s.
A great many books, ranging from learned treatises to globe-
trotters’ notes, have been written on Cambodia, and especially about
Angkor, but IVIiss Ponder’s work cannot be classed with either of these
extremes. A great part of it naturally deals with the history and
antiquities of the country and their gradual discovery by European
explorers and archseologists ; and a long list of sources, besides other
references, indicates that the author has not merely looked at the
venerable buildings but also studied the literature on the subject,
including the latest archaeological discoveries and reports. But she
also had eyes to see, and her own impressions, vividly set down, make
fascinating reading. There is much, therefore, in her work about the
splendid medieval buildings at Angkor and its neighbourhood ; but
there is also much else, for it contains a well informed and sjunpathetic
sketch of the modern state of Cambodia as well. Having paid two
visits to it, with an interval of some years between them, she was
better qualified than the usual hasty traveller ; and her accounts of
1144
REVIEWS OF BOOKS
her travels and experiences, interesting in themselves and illuminated
at times by flashes of humour, illustrate the progress which the country
is making under French guidance. Her chapter on education also
brings out this point, especially in relation to the revival of the old
native arts and crafts, and so does the chapter headed Twentieth
Century Cambodia which deals largely with economic products,
transport, labour, and public health. Other chapters are concerned
with Buddhism and the influence of the monks, law ancient and modern.
Cambodian dancers (of whom a charming account is given), royal
cremations with their costly ceremonial, and many other matters.
Having read every word of the book I can vouch for the fact that it is
extremely readable and interesting.
It may be permi.ssible to refer to a few passages which seem to
be open to criticism. In regard to the statement on p. 102 concerning
the death of Doudart de Lagree, it should be pointed out that though
the expedition of which he was the leailer was in fact the first official
French mission of exploration to ^^sit Angkor, its main object was the
reconnaissance of the Mekong River, with a view to the possibility
of its use for transport, and that his death occurred in Southern
China. It is by no means certain that “ Zabedj '' (p. 114) represents
Java ; probably it was Southern Sumatra. (The word that before
Zabedj is e\ddently a misprint for than.) I see no connection between
the images representing Buddha sheltered by a Naga (p. 181), which
have been found in most Buddhist countries, and the legendary Naga
descent of the old Cambodian kings (pp. 100-1) ; and it seems to me
extremely doubtful that there was a Thai King in Siam in a.d. 607
(p. 291).
I regret to notice a number of misprints, such as Puolo (p. 15, etc.,
for Poulo, the French spelling, or Pido, a current English one), Scanda
(p. 37, for Skanda), Bakahas (p. 50, etc., for Rakahams, as an English
jilural, unless the former is the Cambodian variant), sumptions
(]). 57, for sumptuous), coUonades (p, 57, but on p. 285 with one /),
dieties (p. 62, for deities). Peak Kahn (p. 62, etc., for Prah Khan.
correctly given on pp. 104, 134), v'oman (p. 69, for women), Dangrengs
(p. 77, but the map on p, 157 has Dangrek Mountains), Chantababoiin
(p. 95, for Chantaboun), meterologiral (p. 97), Won (p. Ill, for Won,
i.e. the u dynasty of China), Amarpum (p. 119, for Amarapura).
Tcho-kiang (p. 131, for Tche-kiang. the Chinese province Chehkiang).
htierre (p. 147, for beurrc, unle.ss the first spelling was intentional, to
repre.sent an English nronunciation of tlv French word), Putohitu
BUDDHIST CA\T: TEMPLES OF INDIA
1145
(p. 182, for Purohita), and gimrdien (p. 2G0, etc., ioi ganlien). Some-
what unusual nowadays are pulgeon (p. 17, for pidgin, coupled with
French), Thibet (p. 17) for Tibet, and (so far as my limited experience
goes) the use of garbage (p. 145) in the sense oi garb, or garment. The
appearance of a French or English s (for the plural) in the Khmer
expression pols domreg (p. 202) also seems odd : the nepas-palm
(p. 16) is, I conjecture, the Malay nipah.
It may be an open question whether foreign European spellings
of Oriental names, such as Boroboedoer. Massoiuh (p. 114) and Manoii
(p. 190) should be retained in an English work (apart from verbal
citations in the foreign language). But it is certain that forms like
Icanavarman (p. 111), Crutararman. Crestavarman (p. 113). iaco-
cartnan (p. 114, etc., and on p. 299 Yarcovarman). and Crivijaga
(p. 126), where the French transcription has 9 (not c), can only mislead
the general reader. A spelling like hvarapnra (p. 254). though not
absolutely technically right, is surely preferable. In the nen- faith>^
melted into the neiv (p. 179), one nem .should be old : and open gras>i
land ... is a feature . . . and on them, etc. (p. 212) is a slip.
The reader will be thankful for the maps and plans, although the
former, being on a small scale inevitably give few details. The index
is useful for reference and the illustrations are beautiful.
C. 0. Bl.lgden.
Buddhist Cave Temple.s of India. By M.uor R. S. Mauchope.
pp. ix + 121 with 51 plates. The Calcutta General Printing Co.,
Ltd. 12*-. 6(Z.
No comprehensive account of the Buddhist ca^es of India has
been published since Burgess and Fergusson wrote them up over
half a century ago. Major IVauchope. by condensing their bulky
and inaccessible volumes into 114 not very closely printed pages,
[irovides the public for the first tune with a hand\ guide to those
unique monuments of Indian religiou.s art. Omissions, of course, are
inevitable, and the treatment of the major groups, such as Ajanta and
Ellora, is cursory, but the author brings out most of the salient features
of all the important groups. His survey includes the Udayagiri caves
of Orissa, which are Jain, and the " Brahmanical caves of Elephanta.
Ills introductory chapters and glossary contain soiiie rather crude
mistakes (e.g. form/ ^ ” box. pp. 8 and .^S. gaibha - dome. p. Ill) .
his views on sviiibologv he might \icll ha\e left out . there aie a
1146
REVIEWS OF BOOKS
few blunders in the numbering of the caves and in the “captions "
of his plates. But such blemishes are redeemed by the cleverness
with which he has assembled his illustrations, numbering over one
hundred. Of these, the “ general views ”, which show the caves in
their proper setting, are especially welcome, for previous writers
usually give “ close-ups ” only. The remainder constitute a neat
corpus of types of stupa, column, capital, doorway, fagade, etc.,
which should prove useful to students, as well as sightseers, who wish
to learn the elements of Indian cave architecture.
F. J. K.
Medieval Indian Sculpture in the British Museum. By Rama-
PRASAD Chanda; with an introduction by R. L. Hobson, C.B.
pp. xiv -j- 77, with 24 plates. London, 1936. Price 10s. 6d.
General Charles Stuart, who served in India from 1777 to 1828,
made a hobby of Indian images. When his collection came to auction
in 1872, Victorian England was not interested, there was no bidding,
and the British Museum got it for nothing. But for “ Hindoo ”
Stuart’s enthusiasm, our national collection would be almost as poor
in Indian sculpture as it is in samples of “ Indo-Sumerian ” civilization.
Lord Curzon was the first to awaken Indians to the merits of Indian
art, and it is fitting that, with the zealous co-operation of the museum
staff, an Indian Superintendent of the archaeological collection at
Calcutta, who knows Indian sculpture from A to Z, should “ write
up ” Stuart’s legacy, and the additions which have since accrued.
Mr. Chanda wisely limits his theme to medieval sculpture ;
Gandhara and Amaravati he leaves alone. But medieval sculpture
can only be understood in terms of time and space. On the evidence
of style, subject, and material Mr. Chanda . classifies the collection,
assigmng to each piece its period and province, and he sketches the
evolution of Indian figure sculpture from its beginnings at Bharhut,
Bodh Gaya, and Sanchi.
The bulk of the collection represents the art of the Pala Empire
of Bengal, the early and best phase of which is well in evidence.
There are some interesting fragments from Mathura, a few very fine
examples of the Gupta art of Benares, a number of specimens which
Imk Gupta with Pala, and some rather late and decadent sculptures
from Orissa.
In plain, unaffected language Mr. Chanda explains the differences
POPULAR HINDUISM
1147
between the various “ schools ” of Indian art ; notably the half-
closed eyes of the northern deities and the wide-awake outlook of
those of the south. Students will welcome this concise, well-balanced
introduction to a somewhat difficult subject.
F. J. Richards.
Popular Hinduism. The Religion of the Masses. By L. S. S.
O'jVIallev, C.I.E. pp. viii 246. Cambridge University Press,
1935. 7s. M.
This is a very readable account of modern Hinduism, its
characteristic beliefs and ideals, its worship and ceremonials, the
objects of its worship, its caste system, and its sects. Obviously these
many and varied subjects cannot be dealt with at aU fully in a small
book of less than 250 pages. In fact the main criticism that occurs to
one is that far too many matters are touched upon for any of them
to be really satisfactorily treated. There is also the difficulty caused
by the wide variations of belief and practice in different parts of India.
A reader who has not been in India, and even some who have, may
quite easily fail to reahze that a statement made in one paragraph
about one part of India would be grotesquely untrue of another part
of the country which is dealt with in the next paragraph.
But in spite of this the book gathers together a great deal of
information that it would not be easy to find elsewhere in so small
a compass, and the general impression left on the mmd of the average
reader will probably be on the whole a fairly correct one.
It is a great pity that the author did not use a more exact and
consistent system of transliterating Indian names. There is no
distinction made between Kali and Kali, and the Enghsh reader will
naturally think they are identical. If one writes Parameshvar,
Ganesh, and Vishnu, one certainly should not write Siva and
Salagram and vice versa. On any system of transliteration it is
difficult to see how the spelling Gokhul can be defended.
W. Sutton Page.
Dhola-M.arura Duha. A folk tale in verse from Rajputana. Edited
by Ram SIh, Suraj Karan, and K.arottam D.ls. 8^ X 5^.
pp. 15 + 877. Benares : Nagari Pracharini Sabha, 1934. Rs. 4.
Rajputana is one of the homes of the ballad, and there for centuries
it flourished. Its glory may be dim now, but the past was splendid.
1148
REVIEWS OF BOOKS
and it may flourish once again. In the meantime we must rejoice
that there are men in the country who Io\'e the literature of their
native land and are gi\dng others the pleasure of studying it.
Of the three editors of this poem the first, Ram SIh, is the Director
of Education in Bikaner State ; the second. Siiraj Karan, is the
vice-principal of the Birla College in Jaipur ; the third. Xarottam Das,
is professor of Hindi and Sanskrit in the Dungar College. Bikaner.
The first and second editors have already edited an important
RajputanI work. Prithlraj's masterpiece, Veil Kriscni Rtihunui el.
The ballad, which is 1,348 linos long, is very charming. It has all
the life and \dgour which we are accustomed to associte with such-
poetry ; in fact, it is of outstanding merit even in a country famou.'
for its ballads.
There must bo few languages which have so many cerebrals as
this dialect of Rajasthani. Words like banal, Mdlaranl. halplaj.
chuhnrl, duinnl, duanil, abound on every page ; they give good practice
in pronunciation to European students. The linguistic value of the
poem is indeed very great. RajputanI is not well known, and we
welcome anything that increases our knowledge of it. The mo.st
interesting individual words arc three "‘genitive prepositions’,
^anddu, which occurs five times ; handdu. which occurs twice ; and
hunddu, which occurs once. They remind us of Kashmiri Hand, hmd.
At the end is an index of the first hnes of all dohas and sorathas
preceded by a glossary of about 3.000 words.
This is a book which I have greatly enjoyed. An article on a
subject connected with it is printed in another part of this number
of the Bulletin.
In conclusion may I express the hope that the editors will continue
their fruitful labours.
T. G. B.
Tri.si D.vs's RAMCARmi.I.vAs (R.v.m.vy.xn). Edited with commentary
by R.v:u Xares Trip.vthI. 9| x 7]. pp. 14 + 6 -f -108 + 11-1-
1,291 = 1,6-30. Allahabad : Hindi Handir, 1935. Rs. 5.
It is thirteen years since Shyani Sundar Das published hi.s ('.xcellent
edition of the Raraayan with commentary ; in 1934 appeared the
huge commentary of Janaksutasaran Shltala,sahay Savant (price
3.5), and now we have this one by a well-known poet and writer
YADGAR I VARIS
1149
who has done much to make Hindi and its literature known to the
world.
Ram Naresh Tripathi is the editor of the Kavita KaumudI series.
Four volumes in it are devoted to Hindi, including two (reviewed in
the last number of the Journal) which contain vdllage poetry. He has
also produced a book of agricultural songs collected from Rajputana,
Bihar, and the United Provinces.
This edition of the Ramayan has a long Introduction which gi\ cs
an account of Tulsi Diis's life and works. It i.s pointed out that while
the Ramayan is not a translation from the Sanskrit it is full of echoes
of man}’ (about 200) Sanskrit works.
Of the scores of commentaries on Tulsi Das's Ramayan he nientiou.s
briefly fourteen; he refers also to the English, Oujrati, Bengali.
Bihari, and Uriya translations.
An interesting and useful section is that on Arabic and Persian
words in TulsI's works. All these words are given ; thus the Ram-
caritmanas has 62, while the Kavitavali has 136.
The text is followed is that of the Xagari Pracharini Sabha. The
print is good and clear.
T. G. B.
Y.adgar I Varis. By ZiV-\ ^Ilha.m.mad. 91- .x 7. pp. 207. Lahore,
1935.
The most popular poem in the Paiijab is Varis Shah's Hlr. There
are people who know it by heart. Professional reeiters are in^dted
on special occasions to come and recite it. while large numbers of
listener.s gather to hear tlie words they love so well.
The district.s in which the poem finds greatest favour are those
which lie to the north of Lahore. Shekhupura. Gujranwala. Gujrat.
Jhang. Lyallpur, and Shahpur, the inhabitants of which speak Northern
Panjabi or a t\’po of Lahmli which i.s not very different.
Other poets have written t)f the devotion of Hlr and Ranjha to
each other, but Varis Shah's poem is the best known and best liked.
Mr. Ziya Muhammad has collected all the details he could con-
cerning the life of Vari.s. about which little is known. Unfortunately,
as he himself informs us. most of the storie.s about the poet have no
foundation. In addition to telling us all that could be discovered
about him, he has described his poem, discussing its philosophy,
giving a critical estimate of its poetical value and explaining the
1150
REVIEWS OF BOOKS
reason for its popularity. He has performed a similar service, though
not so fully, for the author's other poems.
Before the time of Vans Panjabi was regarded as incapable of
expressing high thoughts ; the vocabulary was believed to be very
meagre, and when a writer failed to find the word he wanted, he
simply took a Persian or Arabic substitute. But Varis changed these
ideas. He displayed the resources of the language, writing in true
idiomatic style and introducing many of those proverbial sayings
which give such a spice to village talk. Further he showed how a
story should be written, and encouraged his fellow poets to follow
the path he had opened up. In this way he brought new life into the
dejected spirits of his countrymen.
We have good cause to be grateful to Mr. Ziya Muhammad for
the labour he has bestowed on his subject and for the interesting and
useful work which he has produced.
T. G. B.
Nama-lekha aur MuNiBi. By Kastur Mae Bathiya. x 7.
pp. 12 +■ 671. Ajmer: Banthiya and Co,, 1935. Rs. 6.
This is an excellent book by a well-known writer on commercial
subjects. Nearly twenty years ago he brought out Hindi Bahlkhdta,
the first volume of the Saral Hindi Vyapdr ka Granthttmld series.
The third edition appeared in 1927.
Ndma-lekhd is in part a revised and rewritten edition of Bahlkhdtd,
and partly a totally new work. It is divided into two main sections ;
the first has sixteen chapters, the second nineteen. The first deals for
the most part with the same subjects as the earher work ; only the
chapters on hundies, cheques, and promissory notes are reserved
for the second section.
The second part of the book, two-thirds of the whole, is devoted
largely to company law. It discusses the formation and dissolving of
partnerships and companies, describes the laws which govern them,
and gives the details of their books and manangement. Much space
is given to insurance companies, railway companies, factories, and
warehouses, income tax, bankruptcy law, and similar topics.
Mr. Kastur Mai is to be warmly congratulated on the ability and
precision with which he has written : both this book and his earlier one
are worthy introductions to an important branch of science. The
author not only knows his subject, as one might anticipate from his
BHASA EAHASYA 1151
being a Director of the Indian Produce Co., London, and Vice-President
of the Indian Chamber of Commerce in Great Britain, but has
considerable experience in the work of examination, and consequently
is acquainted with the needs and difficulties of students.
There is only one suggestion I would make. A glossary of Hindi
commercial terms and their Enghsh equivalents would be useful to
both Indians and Enghshmen.
The book may be strongly recommended not only to those for
whom it is primarily written, but to those who speak another tongue,
and wish to learn Indian business methods and Hindi business
phraseology.
T. G. B.
Bhas.a Rahasya, Part I. By Shy.Im Sundae Das and Padii
N.Ieayan Acaeya. 9| x 6. pp. 10 -f 106. Allahabad :
Indian Press, 1935. Rs. 4.
In writing this book the authors have had in view the instruction
of their fellow countrymen in linguistic science and in the apphcation
of that science to the languages of India, with special reference, as is
natural, to Aryan languages, above all to Hindi. Much of what is
said will be new to most Indian students, though famihar to Europeans,
especially to those who speak Enghsh, for they have access to it in
English works.
After an introductory chapter the writers discuss speech and
language, types of languages, language families, and Indian languages.
Then comes a long chapter on phonetics which treats of sounds in
general, sound changes, the phonetics of Kharl Boll (Hindustani with
a strongly Hindi tinge), and finally short notes on the pronunciation
of Avesta, Sanskrit, Pah, and Prakrit.
The most interesting part of the book is the section of nineteen
pages on the phonetics of Hindi or Kharl Boli, which is the result
of much study. There are details in it with which I cannot agree, but
I recognize with admiration the enormous labour which the authors
have expended upon it. In connection with this phonetic account of
Hindi sounds attention should be drawn to the Sanskrit-Enghsh and
English-Sanskrit vocabularies of phonetic terms. (There are a few
Hindi words scattered through them.) Students will find that reference
to the vocabularies will, by showing in what sense the terms have been
used, make the description of the sounds much easier to follow.
1152
REVIEWS OF BOOKS
The authors are heartily to be congratulated on their work :
it will open a new world to Hindi-reading Indians, and the pages on
Hindi sounds are of value to Hindi-reading Europeans.
T. Grahame Bailey.
The Jasmike Gaelaxd (Kl\\dam.\l.a). Translated by A. C. Woolxee.
Panjab University Oriental Publications Xo. 27. Oxford University
Press, Humphrey Hilford. 1933. 6.s. net.
Mentioned by writers of the eleventh and following centuries,
but rediscovered only' in recent yeans and first published in 1923,
the K>oida»idla is a modest addition, of uncertain date and author-
•ship, to the dramatic literature based on the last book of the Rdniu-
ijana. Its author's name being given in one IMS. as Diimaga, the Indian
editors hailed the play as the work of Kalidasa's supposed rival (the
Dihnaga of Meghadida 14), whom they further identify with the
Buddhist philosopher of that name. Dr. Woolner finds no good rea.son
for this identification, and on internal evidence holding the play to
be later than the Uttara-Rnhincarita dates it “ somewhere between
the seventh and the eleventh centuries The translation is similar
in execution to that of the Trivandrum plays and its production
equally good.
C. A. R.
The Meghauuta of Kalidasa. Translated from the San.skrit bv
G. H. Rooke. Oxford Univer.sity Press, Humphrey Milford,
London, 1935. 8.s-. Gd. net.
This differs from most verse translations of the famous ly'ric in
appealing to the mind rather than to the ear. The lines, eight to a
stanza, are irregular in length and rhythm. Occasional alliteration
but no rhyme is used. The diction is generally poetical ; where it is
not, the effect to the ear is prose. On the other hand, freedom from
metrical restrictions allows the words to be chosen and placed as
the sense requires, significantly, without padding ; and the whole
has a vigour, ruggedness almost, hitherto unsuspected in Kalidasa
indeed, but full of character and stimulating to the reader's interest.
Opposite to each page of the translation is the text, beautifully'
transliterated into italics, with amusing extracts from Mallinatha's
A SAXSKEIT PRIMER
1153
commentary. The book also has a map of the Cloud's journey, an
introductory poem by Rabindranath Tagore, and coloured capitaLs
at the opening of each part, all excellently printed and produced.
C. A. R.
A Sanskrit Primer. By E. D. Perry. New and revised edition.
Columbia University Press. 1936. 16.s. QJ.
The new edition has a new title-page, a new preface, and a new
price. IVlost of the misprints and errors noted in the Corrigenda of
earlier editions have been corrected, as well as a few not so noted.
Some still remain, e.g. : —
Page 56, sentence 14 : besought . . . were should be beseech . . .
are.
Page 99, footnote : Add or n after dental mute.
Page 188, § 487 ; Add practically before limited.
Page 193, heading : LXV should be XLV.
Page 212, under x/waj : insert pra after -f-.
Page 213, under : 9 aste should be ^asti.
In other respects the book is unchanged : for practical purposes
it is simply a reprint.
C. A. Rylaxds.
Codices Avestici et Pahl.wkt Bibliothecae XXivERSix.xTt.^
H.xfxiensis. Vol. V. The Pahlavi Codex K 43. First Part.
Second Part. With an introduction by Arthur Christexsex.
Ijevin and Munksgaard, 1936.
Both these volumes, containing the two distinct MSS. bound up
together and numbered K 43, maintain the high standard of excellence
of the first four volumes. The texts are as easily legible as the original
as personal use of the MS. enable me to judge. Me have in these
volumes, easily accessible to all. important Pahlavi texts. The fragment
of the Greater Bundahisn and the incomplete Book of the Menok i
Khrat have long been known in Andreas s valuable edition. It is
however very useful to have three complete chapters of the Renkart,
chapters 5, 6, and 9, with parts of chapter 3. Our access to this text
has previously been through two printed editions which cannot serve
as complete substitutes for the MSS. K 43 contains also the 'X’ahman
1154
REVIEWS OF BOOKS
Yast, a Pazand fragment of the Jamasp-namak, the Paragra ceremony,
and part of the Sros Yast.
AH readers of Pahlavi texts have reason to be grateful to Mi
Munksgaard for his publication of these valuable volumes.
H. W. B.
Archeological History of Irax. Bj Erxst E. Herzfeld. Th<
Schweich Lectures of the British Academy, 1934. pp. xi -r H-
London, 1935. 7s. 6d.
These papers represent the three Schweich lectures which we were
privileged to hear in 1934, increased by the additional material which
the time allotted to a lecture did not permit of being included. In
the three studies — of pre-Achsemenid and Achsemenid, of Hellenistic,
and of Sasanian Iran — the author has endeavoured to piece together,
in default of definite written documentary evidence, aU the scattered
data which his many years of familiarity with this field of studies
have brought to his notice. How incomplete this evidence may be is
Dlustrated by the occasional remark that such and such a point is
disputed or doubted. But it was obviously impossible to deal in a
lecture with the doubts of others. Instead we have a brilliant synthesis
of the most various evidence from buildings, legends, coins, inscrip-
tions, and geography. Our old Persian vocabulary is enriched by
the word spadmaida, Bab. madakium “military camp”, which the
author has used to explain the Avestan hammpadmaedaya. The
publication of the full text will be eagerly awaited. One is impressed
by the great use made of architecture as a clue to changing and
developing ideas, and here the excellent plates and frequent illustra-
tions in the text prove most helpful. That in some places the
linguistic arguments suggest doubts is of less moment, but one may
regret to learn that the Khwarna (p. 63) is interpreted as lightning
or thxmderbolt. The whole fascinating story of Gundofarr and the
castle on the Kuh i Khwaja is once again related in the second lecture.
We know the name afostom in P‘austos Biwzandaci (fifth century),
ed, Venice, 1914, p. 333, and in the anonymous Syufiac chronicle
quoted by Hubschmann, Armen. Gram.., 71. These are our earliest
references to the name. They show for the fifth century the disyllabic
form *r6stam. We need not, however, doubt that the form rwtsthm
in Zoroastrian Pahla\d books is old, in use earlier, at least as an
epithet. Written evidence seems still to fail which would bring
HISTOEY or EARLY lEAX
1155
the name Rustam into immediate connection with Gundofarr. It
allows one to suppose that the legend may have been older and
independent of him.
These lecttires sufficiently indicate how much important evidence
has yet to be made generally accessible.
H. AV. B.
History of Early Iran. By George G. Cameron, pp. xvi + 260.
The University of Chicago Press, Chicago, Illinois. 1936. 13s. &d.
The author has recognized and happily sought to fill the gap
which Iranian scholars have always noted : it had not been easy to
find a connected account of events in the land, which later adopted
the name Iran, before the Iranian intrusion into historical documents.
The ideal history of early Elam and its relations with the neighbouring
great powers would doubtless furnish all the relevant historical
documents. The present book is therefore a sketch within limited
space happily supplemented by useful bibliographical notes. After
a brief introductory chapter on the land of Elam and its people, the
history of Elam is traced from the earliest allusions in Babylonian
documents to the downfall of the Median Empire and the rise of
Cyrus 553 b.c. The history is concerned mainly with the pohtical
events. We have, therefore, a valuable chronological framework.
The wealth of royal names, names of pretenders to royalty and of
officials, civil and military, make it a most useful book.
H. W. B.
Les Infinitifs Avestiques. Par E. Benvekiste. pp. IIT. Paris,
Adrien-Maisonneuve. 1935.
In 1904 the Altiranisches Worterbwh of Bartholomae summed
up the results of a vigorous activity in Iranian studies. But researches
have been since then energetically pursued in many quarters, in
Middle Iranian with abundance of new material, but largely also in
Old Iranian. Apart from new views, occasionally an opinion rejected
by Bartholomae has been restored to honour, as here p. 49 piOe
compared with Skt. pitJia-. It may be noted that Pahl. in such
phrases as gardn jyym t hoc viarg “ grievous anguish arising from
death ” {Pahl. Texts, ed. Jamasp-Asana, p. 135, §20), if read plm
1156
REVIEWS OF BOOKS
is probably a derivative of the same verb piyati “ be hostile In
the present volume, M. Benveniste has examined in detail the mam
words at present classed as infinitives. By a stricter definition, indicated
on p. 61, it is shown that many of these cannot rightly be defined
infinitives. The authentic infinitives are then severally treated ot.
In the course of the discussion many valuable interpretations are
offered of difficult passages of the Avesta. A particular treatment is
reserved for the infinitive -dhyai. It is noted that this form is associated
with medio-passive verbs. It is therefore concluded that -dhyai is
itself medio-passive. We seem, however, here to be on debatable
ground. “For the hearing” could mean “that someone hear” or
“ that it be heard ” according as the subject is interpreted. A chapter
is devoted to the Mddle and New Iranian infinitives. The derivation
of -isn from -aOna- is valuable. The suggested derivation of Paste
-dl from *-atai is rendered somewhat difficult by the fact that -il
follows the verbal form which already contains the Old Iranian -ta-
participle ; hral = kr- < krta- and ah A suffix such as -ada might
be more serviceable. We have in Middle Persian amaSisn with
-isn < -adna- added to the -ta- participle.
H. W. B.
Materialien zu den Iranischen Pamirsprachen. Von Hannes
Skold. Worterverzeichnisse von H. Smith. Acta Reg.
Societatis Humaniorum Litterarum Lundensis XXI. pp. viii, 319.
Lund, London (Humphrey Milford), 1936. 25s.
Dr. Hannes Skold was able with the support of the Letterstedt
Fund to visit the Pamirs in 1928 where he toiled at the study of the
dialects of SughnI, Rosanl, Orosorl, Bajul, IskasmI, WakhI, Bartangl,
YazghulamI, and Sahdara. Skold died prematurely before finishing
his edition of this material. G. Morgenstierne, C. Fontaine, H. Smith,
and G. Jarring have collaborated to carry out the work. Skold s
WakhI and IskasmI material will appear elsewhere. In this present
volume we have a most careful edition of the texts in SughnI, Rosanl,
and Bajul with translation. To this is added a comparative list of
words from the dialects studied by Skold, indexes and a SughnI,
Rosanl, and Bajul glossary in which N. Persian equivalents of the
loan words are given but not the etymological equivalents of the
genuine dialect words. The vowel soimds are here distinguished with
SAMDHrNIRMOCANA SUTRA
1157
great acciiracy. In the introduction p. 19 it is even attempted to
distinguish central SughnI, Bajul, Rosanl, Bartangi, and Orosori
on the basis of the pronunciation of the vowel in Old Iran. -an. The
material is a most valuable contribution to Pamir studies. In the
folklore it is interesting to have the zindlrv or zendlry in which the
Av. gandardua survives.
H. W. B.
Samdhintrmocana Sutra. L'explication des mysteres. Texte Tibetain
Mite et traduit par Etienne Lamotte. Universite de Louvain,
Recueil de Travaux publics par les membres des Conferences
d’Histoire et de Philologie 2® Serie, 34® fascicule. 1935. Fr. 75.
The present work is a valuable edition (in transcription) of the
Tibetan text of an early Mahayana sutra of which the original Sanskrit
has perished. With this the Chinese versions have been compared
and the variants quoted. A mosb useful feature of the work is the
presentation of the Sanskrit equivalents of Tibetan phrases, in
particular those with technical meaning. The Sanskritist may thus
easily see what precise Sanskrit word corresponded to the Tibetan
or to its French translation.
The earliest extant translation of part of the text into Chinese
was by Gunabhadra between a.d. 435 and 443. But, as the text
was a source for Asanga whose works are known in Chinese in 413—421,
a date second to third centuries a.d. is reasonably proposed (p. 24—5).
In his preface M. Lamotte has discussed the texts and com-
mentaries, the title, the importance of the text by reason of its early
date and position between the Prajnaparamita texts and those of
the Vijnanavada, and has given a detailed analysis of the contents.
The author has p. 7 himself rightly recognized how little satisfactory
literal renderings of Buddhish technical terms are.
H. W. Bailey.
Akhbar ar-Radi wal-Muttaki from the KiT.iB al-Awr-vk. By
Abu Bakr Muhammad b. Yahya as-Suli. Arabic text, edited
by J. Heyworth Dunne. Subsidized by the E. J. M. Gibb
Memorial Trust. 10 X 7. 308 -h L7 pp. London ; Luzac, 1935.
12s. 6d.
In publishing the section of the Auraq of Es Suli concerning
contemporary poets (reviewed in the Bulletin for 1934, p. 686)
1158
REVIEWS OF BOOKS
Mr. Heyworth Dunne announced his intention of bringing out the
other surviving parts of this important book, and he is making
remarkably good progress. The present part relates to the khahfate
in the last stage of its decline.
Er Radi reigned from 322 (934) to 329 (940), when he was succeeded
by El Muttaqi, who was deposed in 333 (944). At his accession Er
Radi found the dominions of the khalifate much diminished and its
affairs generally in a deplorable condition. Under El Muttaqi httle
territory was left beyond the area in the immediate neighbourhood
of Baghdad and the khalif himself was obliged to wander away from
his capital for some time as a fugitive. After the appointment of Ibn
Ra'iq as amir el umara’ in 324, the khalifs became ciphers in the
hands of their ministers and before long military leaders and other
ambitious subjects contended for the guardianship of their sovereigns.
Civil warfare thus became almost incessant. Ibn Ra'iq, Bajkam.
the Baridis of Basra, and the Hamdanids of Mausil were the most
prominent figures in these disputes. The aim was always the possession
of Baghdad and the disputants in turn succeeded in taking the town
and holding it for one or more periods.
Es Suli's account of the two reigns takes the form of a journal in
which events of varying importance are set down in the order in which
they occurred. It is more detailed and less dry than the ordinary
chronicle to which readers of Arabic history are accustomed. The
author is writing from first hand knowledge and includes personal
experiences and remarks and opinions of his own. He had exceptional
opportunities of observation for his subject. Es Suli had been attached
to the court of the khalifs for years, so that he must have been
acquainted personally with every one of any consequence in Baghdad ;
he had been tutor to Er Radi and when Er Radi became khalif he
appointed him to the small circle of his regular companions, some of
whom were always at hand for his service. El Muttaqi refused to
have any companions of the same description, but Es Suli continued
to take a close interest in public affairs after he had lost his official
employment. He was in Baghdad throughout the time, except for
short absences, and he limits himself almost entirely to events in the
place. When he is obliged to mention something that happened at
a distance, he does so in a summary way, sometimes giving the date
at which the news reached Baghdad rather than that of the incident.
Examples of matters that Es Suli records regularly are the
proceedings of the khalifs, the appointment of ministers and important
AKHBAR AR-EADI WAL-MUTTAKI FROM THE KITAB AL-AM’EAK 1159
functionaries, the arrest and punishment of individuals and the
reasons, the decease of persons of eminence with obituary notices.
Among the miscellaneous items mentioned as they occur, one finds
such things as the movement of troops, attacks on Baghdad, local
disturbances, robberies and remarkable crimes, the price of necessities
in times of scarcity. Es Suli gives his opinion of several prominent
personages and reports talks that he had with some of them. The
one to whom he gives most attention is Er Radi and here he gives the
substance of his conversations fully, not being able as he says, to give
the exact words. Er Radi was an accomplished versifier and Es Sfdi
collects the whole of his poems in the book, consisting of a considerable
number of short pieces totalling some hundreds of lines. Es Suli
also inserts many poems of his own composition, made for particular
occasions, for which he was often called on for a poem.
The ruin of the khalifate was the result of internal disorder rather
than external attack. In Es SuU’s narrative the gradual process
can be followed till near the final breakdown. The loss of revenue in
consequence of the loss of provinces must have caused serious difficulties
to the government, but at first one sees no signs of such a shortage of
money as might have been expected. For instance, the historian
Jahshiyari is called upon to pay no less than 200,000 dinars as a fine,
and Ibn Ra’iq can give 14,000 dinars for a slave girl. A caravan from
Baghdad to Khurasan carries treasure to the ’value of 3,000,000 dinars
and merchandise of equal value. The anarchical conditions and the
grievous damage suffered by the merchants as a result are illustrated
by the caravan referred to being plundered by Kurds and none of
the stolen property being recovered or the robbers being punished.
Another caravan had been cut off similarly by Carmathians a little
while before. Wilful damage to the irrigation canals by belligerents
in the civil warfare was a cause of general impoverishment. Another
was increasing lawlessness at Baghdad where open robbery began to
prevail. In the end such of the inhabitants as were able to lea\ e the
town began to depart. Es Suli tells us how he himself was plundered
by soldiers and made poor for ever afterwards. A striking incident
that he describes is an incredibly barbarous massacre of troops
who had surrendred. Er Radi was by no means devoid of ability
but he was quite unable to cope with the situation in which he found
himself. Among the others no very commanding figure appears.
Es Suli has been disparaged as a historian, because in some of
his books he has been guilty of copying without acknowledgment.
1160
REVIEWS OF BOOKS
Here there is no question of plagiarism ; he is a first rate origin.,
authority, fully justifying El Mas'udi's commendation of his speci.i
merits. It is strange that such a book has had to wait so long fu,
pubhcation. One is glad now to have it before one complete.
Mr. Heyworth Dunne s edition is produced from a unique manuscript
of the thirteenth century, in a difficult hand as appears from s
reproduction given, but tolerably correct, it seems, for the notes (h,
not show many emendations. The editor will have had to supplt
nearly all the pointing, no doubt, and he has vocalized the poetry
completely throughout the book. The difficulties of his task can l,e
imagined. He is to be congratulated on the excellent edition he has
turned out.
In a certain number of passages of the prose the meaning is not
easy to make out and readers would have been grateful for a little
help in the way of punctuation or vowels or for a note where the
reading as printed does not make sense. Some copyist and not the
author is no doubt responsible for El Ikhshid's being called Ahmad, a
mistake that occurs twice, though once he is given his right name
Muhammad. Ibn Hinzaba and El Fa^l are identical, but the inde.\'
treats them as distinct persons. The printing is generally good and
only in a few places does it fail to reach the standard that such a
book deserves.
Rhuvon Guest.
Abou T-Tayyib Al-Motannabi. By R. Blachere. 8vo, pp. .xix A
366. Paris ; Adrien-Maisonneuve. 1935. Fr. 45.
The output of articles and monographs on al-MutannabI has been
so extensive in recent years that Profe.ssor Blachere’s book is doubly
welcome. For not only does he grive a detailed original survey of the
poet s (llwan in its historical setting and literary aspects, but also
a critical summing-up of all these works, Egyptian and Syrian as
well as European. So ample a study deserves a fuller notice, but
we must be content with a brief remark. Admirable as are his analyses
oth of single poems and of the characteristics of successive periods,
one cannot help feeling that Professor Blachere, especially in his
criticism of such pieces as elegies (pp. 106, 212) and his distrust of
e mere y ornamental, sometimes leans towards a “European”
S-hti° '"u'^' J‘’‘^®*®!'™“®®*'^^®*PO“t,intheworkofonewhomhe
ca s a magicien du verbe ”, mannerism becomes affectation
IL CALIFFATO DI HISHAM
1161
is for us an all but impossible task. But in so far as his conclusions
are directed against the misguided appreciations of some recent
Arabic writers, they are justified ; and for the rest, this study, despite
a few bold or even questionable translations, may fitly rank as a
model exercise in hterary criticism.
H. A. R. G.
In Califfato di Hisham. By Fraxcesco Gabrieli. 4". pp. 143.
(Memoires de la Societe Royale d’Archeologie d Ale.xandrie,
Tome VII, 2.) Alexandria, 1935.
Considering that Hisham was the last Caliph to rule over an
undivided Islamic world, it may appear surprising that a reign so
obviously momentous in the history of Islam should have had to
wait till now for monographic treatment. Hitherto V eUhausen s
AraMsche Reich has offered the only critical survey of the material,
but it is noteworthy that Professor Gabrieli confirms his conclusions
on the whole, while revising them in detail. The main feature of this
fresh study is a detailed scrutiny of the government and campaigns
in the outer provinces, which leads up to and supports the argument
that the reign of Hisham was essentially a defensive struggle against
the internal and external forces that menaced the unwieldy body of
the Arab empire, and that to his government was due the miracle
of maintaining its frontiers intact over a period of twenty years .
On the other hand, the deficiencies of our sources are reflected
in the comparatively brief space which is allotted to the internal
administration of the Caliphate. Professor Gabrieli is doubtless
justified in limiting himself to a survey of the direct evidence, and in
dismissing inferential conclusions as “ suggestive e probabile, ma non
documentabile con piii precise prove '. But the question of Hisham s
services in reforming the financial administration in the provinces
must be investigated sooner or later ; in the meantime, the assumption
that such reforms as are hinted at in the sources were merely local
actions initiated by the governors may possibly do less than justice
to the Caliph.
In its discussion of provincial activities. Professor Gabrieli’s work
is a masterpiece of thoroughness, and the present reviewer m partmular
acknowedges his correction of several matters relating to the history
of the Arabs in Transoxania. While the remainder of this notice is
devoted to discussion of some controversial points, these in no way
1162
REVIEWS OF BOOKS
affect this judgment on the execution of the work as a whole. The
argument that Samarqand was recovered by the Arabs in 118736
is not convincing. That imder Asad “ si restaurava pienamente
1 autoiita araba in . . . Transoxiana ” (p. 56) is a statement without
support of any sort ; the only expeditions he led or sent out, so far
as the sources go, were the one in which he attempted to cut off the
water supply of Samarqand, and two into Khuttal. In December,
737, the Tiirgesh army was accompanied by the king of Sughd, along
with the riders of Usrushana and Shash (Tab. 1609, 13). Still more
decisive is the wording of the narrative of Nasr b. Sayyar’s second
and third expeditions in 740 or 741 : tJiumma ghazd ^l-tJidniyata ila
wagharsara wa-samarqanda thumma qafala thumma ghazd ’l-thdlithata
ila l-shdsh (Tab. 1689, 7-8). Can ghazd ild in such a context imply
anything hut an expedition into hostile territory ? In regard to the
supposed execution of Kursul on the Shash expedition, it is surely
evident that the second narrative (from an Axdite source) in Tab.
1691, foot-1692 is a parallel or variant to the first (1689-1691, anony-
mously from al-Mada’inI), and this contains no indication that the
Tiirgesh were concerned in the matter at all — which is, for the rest,
unlikely in view of the internal troubles of the Tiirgesh after the
assassination of Su-Lu. Besides, too many good reasons have been
put forward for the identification of Kursul with Baga Tarkhan
for them to be simply set aside by accepting at their face value the
typically romantic ” features of the first narrative. On p. 46 and
n. 1, Khusraw was certainly either son or grandson of Peroz, since
the Chinese documents (Chavannes 172, 257-8 ; cf. Arab Conquests,
p. 16) make it quite clear that Peroz was the son of Yazdigird and that
he was long since dead. In the passage quoted in n. 2 on the same
page, Farghana may perhaps be an oral corruption of Farawa or
Faghandiz, both villages in the vicinity of Bukhara. Finally, it
still appears to the reviewer not impossible that the traditional
account of the measures of al-Ashras may reflect an attempt to
introduce the separate kharaj and jizya system (not a “ direct inter-
vention of Hisham, as Professor Gabrieli has misunderstood it,
p. 44, n. 4). The text quoted in n. 1 on p. 42 shows that in 728 these
terms were stiU undifferentiated in Khurasan, and the subsequent
troubles may well have arisen from a misunderstanding due to this
confusion of meaning.
H. A. E. G.
G. BEBGSTRASSEr's GRUXDZt^E DES ISLAMISCHEN EECHTS 1163
G. Bergstrasser’s Grundzuge DES IsLAMiscHEN Rechts. Bear-
beitet von Joseph Schacht. 8vo, pp. viii + 145. Berlin and
Leipzig ; de Gruyter. 1935. RM. 12.
This compact work has a double object. One is directed to the
student of comparative jurisprudence, who is given an analysis in
juristic terms of a standard work of matured Hanafi Law (the Multaqa
’l-Abhur of Ibrahim al-Halabi, d. 1549). The other is addressed
rather to the student of I.slamic culture, and sets out the principles
which underlie the formulation of Islamic Law and their place in
the general field of Islamic thought. The title reflects this duality ;
to the former it is only the last four words that matter, while its
main appeal to the Orientalist is contained in the first three. The
width of Bergstrasser’s scholarship, his depth of learning and scientific
honesty lend special authority to his views, and we owe a debt of
gratitude to Professor Schacht and Professor Pretzl for reseuing these
lectures and notes from his MS. materials.
The introductory chapters contain — apart from the initial
generalization — little that is not generally accepted. Then follows
a valuable analysis of the basic concepts applied to legal transactions ;
intention, statement, stipulation, suspension, representation, and
especially validity and nulhty. The rulings themselves are grouped
with remarkable concision in eight chapters, and the book ends with
a short chapter on the limitations to the apphcation of the Sharl'a
and an original discussion of its peculiar nature and structure.
While the general substance and practical value of the book are
above criticism, it poses, for the Orientalist, two main questions.
Is Islamic Law, as the opening sentence asserts, “ the epitome of the
true Islamic spirit, the most decisive expression of Islamic thought,
the essential kernel of Islam ? ” Secondly, is it possible to restate
the contents of the Sharpa in current juristic terms without some
degree of deformation ? The author himself seems to answer the first
question in the negative : das Recht existiert nur als Bestandteil
der ganz anders orientierten religiosen Pflichtenlehre (p. 123), and
with it he indicates also the answer to the second, about which, to
do him justice, he expresses some uneasiness. Only by a forcible
abstraction of Law from its ethical foundations (and this imphes also
its social environment) can it be so treated. As Professor Schacht
has himself indicated elsewhere (in OLZ., October, 1935), the scholastic
formulation of the Shari‘a attracted Bergstrasser more than the
problems of its application and meaning in the Islamic community ,
1164
EE VIEWS OF BOOKS
and it is only this concentration upon the outer forms rather than
on the spirit which can account for the astounding statement on
p. 115 that the object of the qadi in the hearing of a case “ is not to
find the truth . . . but only to observe given rules Such an assertion
makes one wonder whether Bergstrasser ever asked himself the most
fundamental question of all ; what is the meaning of justice ” to
the Mush'm ? And it is noteworthy that the word insaf, which most
characteristically sums it up, is never mentioned in this book.
H. A. K. G.
Revue des Etudes Isl.uiiques. Tome IX (Annee 1935). 4 cahiers,
pp. 431. Paris ; Geuthner. Subscription price 100 francs.
Of the fifteen articles in this volume, the longest and most
important is a systematic survey of the details relating to rehgious,
social, and administrative affairs in Tunisia in the early Islamic
centuries to be gleaned from a fourth-century biographical work,
Rirjad an-Au/us (105—177, 273-305). The author, H. R. Idns, shows
what a rich harvest awaits the student who has the patience to winnow
a section of these often uninviting materials, and sets an excellent
example of how to do it. A curious appendix to his survey is furnished
by T. Lewicki, on the survival of ram-worship in Tunisia down to
the third century a.h. (195-200). Another historical article, by A. N.
Pohak, though interesting and original, is less convincing (231-248).
In seeking to demonstrate the “ colonial character ” of the Mamluk
state in relation to the Golden Horde, the argument for the “ vassal
status of the former is pushed too far, especially when the main
features of its organization are ascribed to Mongol influence.
Incidentally, creoles ” seems a Uttle crude as a rendering of awlad
an-nds. The series of Abstracta Islamica is continued by two hands ;
P. Kraus deals critically with philosophical studies (A 217-238),
but H. Charles gives httle more than a list of works published in
other fields (A 239-293).
The other articles deal with present-day problems. J. Beyries
contributes two valuable studies of the rehgious schools and the
new Ghudf fraternity in Mauritania (39-73) ; P. Marty, under the
unpromising title of “ L’aimee liturgique musulmane a Tunis ” (1-38)
gives a mass of interesting notes on the festivals and amusements of
the population ; and J. Schombart de Lauwe briefly surveys the
native trades unions in Algeria and Tunisia (187-194), all languishing
BEITRAGE ZUR ISLAillSCHEX ATOMENLEHRE
1165
at present but capable of being rapidlv' transformed into dangerous
political associations The laicization of justice and Muslim reactions
to it in Albania are briefly sketched by G.-H. Bousquet (399-410),
and Professor Masse contributes a note on the unveiling of Persian
women, mainly in Teheran (411-18). Aziz Bey Hanki cites recent
judgments in Egypt determining the right of non-Muslims of foreign
residence to inherit from resident relatives (179-186). Analyses are
given of two recent publications which caused some controversy at
the time of their issue : the decree condemning the monograph of
‘All ‘Abd ar-Raziq. translated in the two previous volumes, and the
feminist work of the Tunisian Tahir al-Haddad (75-86, 201-230).
The vexed question of Latin alphabets is raised by P. Rondot in
discussing the alphabet adopted by the Kurds in Russian territory
(87-96), and by a scheme for the transliteration of the Arabic alphabet
by Kafiz Danishman (97-103), which is perhaps too comphcated but
has some advantages over any hitherto suggested.
Finally, Professor Zeki ^'alidi surveys the work of a number of
Oriental scholars who received a European training, and makes a
plea, to which most Orientalists would gladly assent, for closer
collaboration between Oriental and European students of the East
(249-271), But whether scholars in the Arabic countries would wholly
agree with his presuppositions is doubtful ; and it is perhaps significant
that he misquotes Dr. Taha Husain when he represents him as saying
that the student must choose between the old Oriental and modern
European methods ; what the latter actually says is : ‘‘ Both methods
are indispensable if we wish to make a thorough study of Arabic
literature.”
H. A. R. Gibb.
Beitrage zitr Islamischen Atomenlehre. By Dr. Salomon Pines.
pp. 147. Berlin, 1936. RM. 4.
The Muslim theory that there is no order of nature, that God creates
the effect with the cause, is well known. In this book Dr. Pines studies
the beginnings of this theory in the theologians of the second and
third centuries of Islam. His w’ork is an advance on that of the late
Max Horten, if this can be said without ingratitude to the dead. He
does not assume that the theologians knew the most recondite problems
of Greek philosophv. Perhaps he is still too intellectual. The doctrine
of Abu T-Hudhail that there is nothing doing in heaven, is an attempt
1166
REVIEWS OF BOOKS
to describe a place of rest and peace, he might have used the words
of Theodore Abu Kurra “ when the resurrection takes place and all
things become unchangeable ” : it is not an apphcation of the principle
that what has a beginning must also have an end. The author calls
attention to the influence of the Koran on terminology and to the
variety of thought during the formative period, a time of intense
rehgious activity. One complaint. Why Mu‘ammar ? The Kitcib
al-Ansdb has Ma‘mar.
At that time everything in the world was either substance or
accident. By a.h. 200 the atomic structure of substance was accepted
by most as also the impermanence of some accidents. A little later
it is clear that time was conceived as a succession of units. Probably
before 200 al-Najjar taught that capacity, which is not the same as
power, existed only for a moment. It is odd that the later theologians
do not draw nearer to orthodoxy on this question as they do in others.
The author does not notice two subjects of discussion, whether an
accident can be renewed and whether one unit of power can do two
units of work. The second part of the book deals with Muhammad
b. Zakariyya al-Eazi, the arch-heretic. Time and space have now
become subjects of discussion. In his doctrines, in opposition to the
ruling Aristotehanism, appear ideas which were believed to be
Platonic. The third part deals with Indian philosophy and it is made
probable that the Muslim atomic theory came from India. Dr. Pines
has written a thorough and sane book ; if his conclusions are imcertam,
that is not his fault but the fault of his material. We hope that he
will continue to pubh'sh the results of his studies.
A. S. T.
Die Pbophetie : 0 Istocie Prokoctwa. By A. Heschel. pp. 194.
Krakow, 1936.
This book is published by the Oriental Section of the Pohsh
Academy and is written in German with a summary in Polish.
What is a prophet ? This is the question which the author sets
out to answer by studying the pre-exihc prophets, omitting all that
is not directly connected with the fundamental theme. His answer
is that a man becomes a prophet when he has to announce that God
is not indifferent to what men do but feels keenly their treatment of
Him. Amos proclaimed the sorrowful disgust of God at the wickedness
TKAITilS EELIGIEUX, PHILOSOPHIQUES ET ilORAUX 1167
of men. Hosea taught that this disgust is caused by God’s love for
His people, a love which has not lost hope. Isaiah is more concerned
with the manifestations of God’s feelings and lays stress on His
disappointment with men s obstinacy in evil doing. Jeremiah lays
stress on the outraged love of God which makes Him resolve to punish
sin ; His anger is not His real nature but the reply to sin. The secret
of rehgion, then, is the suffering of God. In the second part of the
book this theme is followed through theologians and philosophers ;
curiously enough, the author does not refer to the New Testament
where it supports him, where it is the direct heir of the Old Testament
prophets, though he does refer to it and to Christian theologians when
he does not agree with them. Unfortunately, the book is written in
the worst German style, pretentious and in places incomprehensible.
When the author argues from first principles that prophecy is not
ecstatic, his method is scholastic not scientific.
A. S. T.
Teaites Religieux, Philosophiques ET iloRAux, Extraits des
(Euvres d’Isaac de Ninive. Edited by R. P. Paul Sbath.
pp. 128. Cairo : Al-Shark, 1934.
Isaac made a selection of golden thoughts from the works of
Chrysostom ; years later he himself suffered the same fate. However
good it may be, morality in tabloid form is cloying. The sayings of
Isaac were translated into Arabic and the editor has corrected the
language to bring it into line with the grammars but as he gives the
original in the notes, we can still see how a Christian wrote Arabic
in the ninth century. There are a few mistakes ; thus the cathohcus
Yuhanna was the son of Narsi not of Barsi. The pamphlet is interesting
for certain resemblances to Mushm thought. That the acts of men
do not arouse either anger or pleasure in God is close to the Mushm
belief that men can neither help nor hurt Him. That intention makes
a good act is sound Muslim doctrine. The discussion on man's share
of worldly goods (rkk) is like that on the possibihty of things tabu
being part of God’s provision for man. One would like to know who
were the Christians who taught that God's will was not eternal, a good
Mu'tazili doctrine. The book has interests beside those of devotion.
A. S. Trittox.
1168
KEVIEWS OF BOOKS
A^r Early Mystic of Baghdad. A study of the life and teaching <
Harith b. Asad al-Muhasibi. By Margaret Smith. x 5
pp. xi + 311. London : The Sheldon Press, 1935. 15s.
It is astonishing that no monograph has hitherto been writtt
on Muhasibr : still more astonishing that none of his very considerab!
bulk of writings should have been edited, with the exception o
Dr. Ritter’s text of the Bad’ man andba (for of the editions of tw-
tracts mentioned by Miss Smith on p. 57 of the present book I find ra
trace elsewhere). Of the importance of Muhasibi and his profouni'
influence on the systematic development of Sufi doctrine there car
be no doubt : yet Massignon was the first to call serious attention tr
the fact, and Miss Smith is the first to investigate the matter in detail.
Miss Smith’s book is full of materials otherwise unpublished, and
perhaps its chief drawback is that it attempts to deal with too many
things at once. One could have wished, for example, for a separate
detailed analysis of the Ri'dyah : even more desirable would have
been an edition of that work preceding the present study, that one
might have had materials for reference. Miss Smith says, however,
that such an edition is under way, and one hopes that it will see the
light before long. It is abundantly clear that the Ri‘ayah is the most
important single work of Sufi authorship before Ghazali’s Ihya •
To Miss Smith belongs the credit of having proved beyond all doubt
the direct and often verbal indebtedness of Ghazall to Muhasibi.
This is the most valuable fact which emerges from the present book,
and her article on the same subject in the JRAS. for January, 1936.
To criticize a work so discursive and based to such a great extent
on manuscript materials is impossible within the scope of a review.
It may, however, be permitted to remark on a few transparent defects.
Ghazall’s “ Deliverer from Error ” always appears as Munkidh
(instead of Munqidh), a shp to which Miss Smith appears to be wedded,
for it reappears in her article referred to above. Jllanl is credited
(p. 257) with a Ghurriyat (i.e. Ghunyah), and the index repeats the
ascription ! Greater care is absolutely indispensable in scientific
works. I have been able to collate the translation (pp. 18-20) of the
exordium of the Nasaih, and have detected in it no fewer than six
instances of faulty rendering.
Such defects as these should not, however, be allowed to obscure
the fact that here Miss Smith has written a book which is of its kind
one of the most interesting and valuable contributions to the study
of Sufism in our language.
A. J. Arberry.
THE DOCTRINE OF THE SUFIS
1169
The Doctrine of the Sufis. Translated from the ^li’abic of Abu
Bakr al-KalabaAl. By A. J. Arberry. pp. xviii + 173. Cam-
bridge University Press, 1935. 10s. Qd.
This is a translation of the Kitdb al-Ta‘arnif li maMhab ahl al-
tasau'iviif, of which the translator published an edition last year.
The writer, Kalabadhi, died towards the end of the tenth century a.d.,
and the work is known to students of Sufism as one of the oldest
treatises on the subject. Although it is not so full and valuable an
exposition of Sufism as the Qut al-Qidub of Abu Talib al-MakkI,
Kalahari’s contemporary, or so clear and systematic as the Risdla
of al-QushayrT,i and the Kashf nl-Mahjub of the Persian Hujwlri,
it is of considerable interest because of its early date and the sayings
and anecdotes of the Sufis which it has preserved to us.
After a number of disconnected chapters dealing with the views
held by Sufis on a variety of subjects — including gnosis (Ai as
distinguished from knowledge (^), of which a Sfifi said, “ God has
made knowledge free to all believers, but gnosis He has reserved for
His saints ” — Kalahari proceeds to a more systematic consideration
of the Sufi way of life and its goal. He deals very briefly with such
stations as repentance, asceticism or renunciation (a better translation
for Jjtj than “abstinence” which the Sufis commonly express by
^Jj), patience, poverty, humility, piety {^^), sincerity, or “ single-
mindedness ” — which comes nearer to the full significance of
gratitude, trust, satisfaction, and certainty which is the attain-
ment of knowledge of spiritual things, and the setting aside of the
veils that hide it from the seeker.
In his account of Concentration Kalahari uses the very
words and phrases of al-Muhasibi, who, a century and a half before,
had taught the need for the unification of all the powers in con-
centration upon the One, so that the seeker’s concern might be simplified
and unified and become one single concern (b^[} ^)-
Kalabadhl writes at greater length upon bi, the death to self, and
the fife in God (for which “ persistence ” seems a very inadequate
term). Of the mvstic who has entered into the life lived in, and
through God, the author says that “ what he does, he does unto
‘ Dr. Wilson Hume, of Lahore, is now working on a translation of this treatise.
1170
REVIEWS OF BOOKS
God, not seeking any joy for himself, either in this world or the next
— he has passed away from his own attributes and subsists in those
of God The sign that the mystic has died unto self, is that he has
ceased to desire this world or the next, his desire is unto God alone,
and so he attains to the vision of God Alone in His Unicity. Now he
has become altogether receptive, open to the Divine gifts, and, as
Junayd taught, he becomes that which God meant him to be, and he
returns to the state in which he was at the first, before he came into
bodily existence (pp. 123, 124). This is a state which the great Sufis
held to be permanent, and henceforth the mystic fives the mutive
fife in God.
We should judge that Kalahari was not himself a mystic, and
for this reason his exposition of mystical doctrine, in its cautious
orthodoxy and its brevity, falls short of the account of writers who
were themselves mystics ; who had, therefore, a fuller comprehension
of the subject and could supplement the teaching of others by what
was derived from their own experience.
The translator has aimed, admittedly, at a literal translation of
the text, and this, no doubt, accounts for some sacrifice of literary
form, but there are signs of over-hasty production or lack of revision
in the inclusion of such sentences as “ He commands us to be obedient
much ” (p. 91), “ They faint of hunger ’’ (p. 6), “ He has never ceased
thus ” (p. 16), and the translation includes phrases which are un-
necessarily clumsy and obscure. “ Disassociation from the soul ”
(p. 90) would be better rendered as “ relinquishment of the self ’
(i.e. self-surrender), since is nearly always used by the Sufis
in the sense of the self, the lower nature. By no effort could a man
“ disassociate ” himself from his soul. The examples of Sufi poems
included by Kalahari have been translated into English verse and
certain of these are very attractive.
The book has been well produced by the Cambridge Press and has
an adequate index.
Margaret Smith.
Three Treatises on Mysticism. By Shihabuddin Suhrawerdi
IVIaqtul. Edited and translated by Otto Spies and S. K. Khatak.
Bonner Orientalistische Studien. Stuttgart, 1935.
This edition and translation of three hitherto unpublished Persian
treatises by Shihab al-Din Suhrawardi al-Maqtul, which includes a
Persian commentary on the last of the three, and also a Persian
THREE TREATISES ON MYSTICISM
1171
biography of Suhrawardi by Shahrazuri, will be welcomed by students
of Islamic mysticism.
The author of these treatises, Suhrawardi al-Maqtul was put to
death by the order of Salah al-Dln (Saladin) in a.d. 1191, when he
was but thirty-six years of age. He was an able thinker, who called
himself “ The Seeker of the Invisible World ’’ (al-Murld bil-Malakut).
and became deeply absorbed in the study of Sufism. In his most
characteristic work, the Hikmat al-Ishraq, he claims as his forerunners
the Greek philosophers, Agathodaemon, Hermes, Empedocles, Pytha-
goras, Socrates, Plato, and Aristotle, and also the Persian sages
Jamasp and Buzurgmihr. Suhrawardi himself was a pantheist, who
developed an interesting doctrine of illumination, holding that God
was the Light of Lights (Aar al-anicar) which brought all things
into existence, by pouring out its rays into their being. The human
soul, partaking thus of the Divine, seeks to return whence it came
forth, and when, by the purification of self-discipline, it has removed
the veils between itself and the Primal Light, it attains to the Beatific
Vision and to union with the Divine.
The three mystic-philosophical treatises included in this volume
are the Liighat-i-Murdn (The Language of the Ants), the Safir-i~
Simurgh (The Note of the Simurgh), and the Risdlat al-T ayr (Treatise
of the Bird), the latter being a Persian translation from the Arabic
treatise of that name, by Ibn Sina (Avicenna).
The mystic Simurgh, which gives its name to the second of these
treatises, and was also the subject of ‘Attar s Mantiq al~Tayr, is indeed
a symbol of the One and All, and of it Suhrawardi writes, “ Know
you, that all colours are in him, but he is colourless, and his nest is
in the East, and the West is also not devoid of him. All are occupied
with him and he is free from all. All are filled with him and he is
empty of all ” (p. 29). This treatise is, in fact, concerned with ma‘nfa,
the gnosis which is the intuitive knowledge of the One and AH. Like
Qushayrl and al-Ghazall, Suhrawardi writes of the Divine revelation
as coming at first as brief flashes of light, and then as remaining,
which is the state of Tranquillity (sakJna). From tranqmlhty the
mystic passes into that state of annihilation of all personal attributes
‘ Cf. Qushayrl, " It begins with flashes of light, then rays, then the light shining
forth in its full splendour.’ Jiisala, p. 53 (ed. Cairo. 1867) and Ghazali, ' If the m>s le
IS free from all distraction, the Light of God will shine upon his heart and it will be
at the first like a blinding flash of lightning. Its sojourn is but brief, but it will return,
for it IS the prelude to a constant communion with God. Ihyu , lii, » i
a.H. 1272).
VOL. vni. PART 4.
75
1172
REVIEWS OF BOOKS
which is called fund dar fand. Unification (taivJud) he divides inti
five grades, of which the highest is the stage of those who lose them-
selves in God, and beyond that there is no further stage, for it is tL
entrance into Eternal Life (p. 38).
This, too, though the editors do not note it, is the teaching n:
Ibn Sina given in his Fl maqdmdt al-‘Arifin (Stages of the Contem-
plative Life), and Suhrawardi seems to have derived much of thn
second treatise also from Ibn Sina.
In his teaching on man’s love of God, Suhrawardi points out that
man’s highest happiness consists in the attainment of perfection,
whether it be in regard to the natural or the moral faculties, e.g. the
eye delights in perfect sight and the ear in perfect hearing. So the
highest happiness of the soul consists in the attainment of its perfection,
which is the intuitive knowledge of Truth and the comprehension of
ultimate realities, for “ the human soul is the noblest of the Seekers
and Truth is the greatest of the “ known
The translation, on every page, bears witness to the fact that
English is not the mother-tongue of the translator, and, while it
gives an adequate interpretation of the meaning of the treatises as a
whole, the translation of individual words and phrases leaves much
to be desired, and it is necessary, at times, to refer to the Persian te.vt,
in order to be sure of the author’s meaning, p. 19, 1. 5 from foot
(translation) should be “ They threatened to make it look at the
sun ”, rather than “ They terrified it of looking at the sun ”, p. 4L
1. 7 from top, should read “ it is not fitting that man should take God
for his friend ”, and, on the same page, 1. 9 from foot would be better
translated “ Passion {‘ushq) is to be defined as love which has been
carried to an extreme ”, and the last line of that page as “ Yearning
(shawq) is defective because ‘ non-acquirement ’ is an essential
part of it ’ , i.e. it has not yet found what it seeks.
Among misprints to be noted are evistant (p. 31), dovotee (p. 33),
and beleive (p. 51). The book is well printed, but the complete lack of
an index is a serious defect.
Margaret Smith.
Tarikh-i pansad sala-yi Khuzistan. By Sayyid Ahmad Kasravx.
pp. 287. Tehran 1313/1934. London Agents : Luzac and Co.
6s. 6d.
S. A. Kasravl is known to the readers principally by the three
volumes of his very interesting essays on the “ forgotten ” dynasties
1173
taeTkh-i pansad sala-yi khuzistax
of North-Western Persia. His new work on the rulers of the south-
western corner of Iran merits our full attention both by the skill
with which the subject is treated and by the profusion of new materials.
The book begins with the advent of the d}'nasty of Say\nd
Muhammad b. Falah known under the name of Musha‘sha'. This
founder of an extremist shf a sect was born in IVasit but, after several
unsuccessful risings of his followers in the region of Mesopotamian
marshes, he occupied Hawiza on the lower course of the Karkha and
here his descendants ruled first independently (a.d. 1458-1508), then
as icatis of “ ‘Arabistan ”, on behalf of the Safavids (a.d. 1508-1722),
and finally as obscure local rulers eclipsed by the rising star of the
Ka‘b (Cha‘b) chiefs.
It is a matter of regret that the Persian historian disregarded the
existence of two very thorough articles on the Musha'sha‘ by Professor
W. Caskel, in Islamica, iv, 1, 1929, and vi, 4, 1934, who examined
all the sources accessible in Europe and, among them, even such rarer
works as the qasidas of the paneg^Tist Ibn Ma’tuq (d. 1007,1676).
S. A. Kasrawi necessarily uses much the same material but he
adds to it many new sources. Above all, he discovered in Zanjan a
copy of the original Kalam al-Mahdi written by S. Muhammad b.
Falah in imitation of the Qor'an and containing very important data
on his religious and political theories. Several of the suras of this
Kaldm are quoted in the Appendix of the book, though in the text
S. A. Kasravi's presentation of these doctrines is a little obscured by
his own orthodoxy. On p. 26 he suggests that heretics similar to
S. Muhammad “ ought to be fearlessly exterminated like dogs .
For the author’s personal ideas the passage on p. 53 is interesting,
in which he distinctly disapproves of the Safavid rebgion. Speaking
of Isma‘I] II, who was notorious for his Sunni leanings ^ he says .
” unlike the other Safavids, Ismii'il was not trammelled by religious
innovations (bid‘at), and seems to have striven to remove the bad
practices {zisht-hdrihd) introduced by his grandfather and father ,
this was the reason why he acquired the reputation of a sunni {siinni
<jart).”
Another book used by the author for the first time is the history
of the Musha‘sha‘ written by one of themselves, Sayyid Ali b. S.
‘Abdullah b. S. ‘All-khan, wldch brings the events down to the times
of Nadir-shah (pp. 7, 97).
The author quotes several other rare works like the TaJcnnlat
^ See now Hinz's special article on this king in 4/XOS., 1933. pp. 19 100.
1174
EEVIEWS OF BOOKS
al-alchbar by ‘Ali b. ‘Abd al-Mu’min, a contemporary of Shah Tahma- >
(p. 48), and a number of tribe chronicles, official documents, memoii
etc. Unfortunately the author nowhere describes these interestii
sources systematically and much time is needed to ascertain tl,
identity of a source quoted. So with regard to the Ka‘b (Cha n
tribe the author quotes a Daftarcha-yi tarikh-i Kab beginning wit
A.H. 1106 (pp. 119, 148), which he also calls Tdnkh-i Ka‘b (pp. lb
126, 149). Nothing v'ery clear is said about the Maqaldt of Shayk;
Fath-allah Ka’b (pp. 142, 146, 148), identical, as it appears, wit'
Kaffir’s 7,ad al-musafir (pp. 69-71). The author uses an epitome ot
the books called Riydd al-‘ulamd and Tuhfat al-Azhdr (p. 8) withoui
saying a word about their authors. This absence of method is irritating
in a serious historical work.
The title of the Musha‘sha‘ is still not very clear in spite of M
Caskel’s special note, Islamica, 1929, pp. 91-2. In Dozy's Supplement
i, 764, is explained as “ rayonner ” and as ‘‘ radiation.
irradiation . . . en parlant de tuiles d’or et d’argent ; aussi en parlant
du vin qui, lorsqu’on en boit beaucoup, rougit le visage et le fait
rayonner In the Kaldm al-Mahdi (Kasravi, pp. 274-5) the founder
of the dynasty applies the term sha‘sha {*sha‘sha‘a) to the two
principal episodes of his earlier career, namely the rising among the
Ma'dan tribe (read ; instead of and that in the locality
Dub, cf. Majdlis al-mu’mimn. Consequently sha‘sha‘a practically
means something like “ upheaval ” but, remembering Dozy’s inter-
pretation, we may take it for “ (mystic) inebriation [‘ the wine going
to the head ecstasy ”. The author of the Majdlis confirms that
Muhammad b. Falah taught his followers a dhikr which produced
a state of tasha‘shu‘ (“ ecstasy ? ”) in which they became capable of
astonishing feats : walked over burning coals, etc.
One of the later walls of ‘Arabistan is mentioned by many con-
temporary authorities as the traitor who facilitated the capture of
Isfahan by the Afghans, which was followed by the deposition of
the Safavids in a.d. 1722. S. A. Kasravi (p. 102) takes him for
Muhammad b. ‘AbduUah, but a Dutch witness has preserved for us
the name of the villain ; he was called ‘Abdullah (son of Farajullah
and father of the aforesaid Muhammad ?), and consequently was the
man whose adventmrous career is described by Kasravi on pp. 98-101,
see Alexander a Sigismondo’s original report translated into English
by H. Dunlop in J .R. Central Asian Society, October, 1936, pp. 647-8-
IL-FIDWA TAL-BDIEWA
1175
Apart from the Musha‘sha% S. A. Kasravi treats in detail the
history of the Ka‘b established in Khuzistan, and finally of the chiefs
of the Muhaysin clan, whose last representative was the well known
Shaykh Khaz'al. The story of the latter's attempt to oppose in 1924-5
the present Shah of Iran (at that time generahssimo of Persian troops)
is told by the author as an eye-witness. Shaykh Khaz'al was deported
to Tehran and died in exile on 25th May, 1936.
On the whole S. A. Kasravi’s book is a very welcome addition to
the modern historical hterature on Persia.
V. Minorsky.
Il-FiDWA TAL-BDIEWA. DrAMM b’HAMES TAQSIMIET U DeHRA.
By A. Cremona, pp. 171. Sm. 8vo. Malta : Empire Press, 1936.
In the Bulletin for 1935 (pp. 173 seq.) we were told of the vicissitudes
of the theatre in Egypt and of the opposition on religious grounds of
some of the stricter Mushms. Malta has been rather more fortunate
in this respect, for it has had a theatre of its own since 1732, when
Grandmaster Manoel de ViLhena opened the fine “ Manoel Theatre ”,
which still exists, though now mostly used for the “ pictures ”. But
that theatre was intended for drama and opera produced in foreign
languages for the benefit of friends of the knights and those who had
assimilated their culture. The idea was a good one but needless to say
the masses were not touched by it. But the theatre is too good a
medium of education as well as of recreation to be neglected by those
who had progress at heart. After the departure of the Knights, the
native influence again began to assert itself, for the playfully satirical
side of the Maltese character is not easily suppressed. About 1830
we hear that G. P. Badger, afterwards so well known for his Lexicon
but then a mere lad, son of a soldier of the garrison, was organizing
dramatic entertainments in the vernacular for boys of his own age.
Efforts on these lines continued, but never got into print. The earliest
native comedies known to me are two by Carm. Camilleri printed in
1860 and 1877. Since then other pieces have been published, notably
by M. A. Borg.
The present work is something rather more ambitious than any
thing I have seen before. It is a rural tragedy in five acts, with a
cast of fourteen actors and eight actresses, besides supernumeraries.
It is in eleven-syUable blank verse, with occasional rhymes. Its
setting is in the early fifteenth century, in the Aragonese period.
1176
REVIEWS OF BOOKS
The title of the play (The Redemption of the Peasants) shows tL
subject. The King had pawned the Island of ilalta to his Sicilian
Viceroy, who in turn had pledged it to 3Ionroy. The Maltese thereupon
determined to “ redeem their country by repaying the amount of
the original loan and so becoming direct lieges of the King.
C. L. D.
Ugo Foscolo. L-Oqbra. Dahla, qlib u tifsir, bi ^jieda ta‘ hajj \
QIL-QOSOR ta‘ Ugo Foscolo. By Ultn Karm. pp. 65. Sm. 4n
Malta : Government Press, 1936. Is.
In 1927 Italy kept the centenary of Foscolo’s death and the
thought then came to “ Dun Karm ” (Mgr. Psaila) of translating
into Maltese the best known and most difficult of that poet’s works
I Sepolcri Two years later there fell, quite unnoticed, the centenary
of the death of Mich. .:\nt. Vassalli, the first to bring out a Maltese
grammar and lexicon. A certain likeness between the fives and
aspirations of the two men, both of whom spent long vears in exile
for political reasons, and both of whom, the Italian and the Maltese,
were devoted to their native lands, set Dun Karm to work, and the
present elegantly printed book is the result.
Foscolo s work, by contrast rather than by likeness, reminds one
of Geo. Eliots “ 0 may I join the choir invisible With Gray's
Elegy the coimection is very remote. Dun Karm’s translation
is rather a (our de force and clearly involved much labour. The
translation is in blank verse like the original, and carries eleven syllables
to the fine. The copious and informative footnotes explain difficulties
in the text and justify the translation. In the title, Oqbra is the plural
of qabar (tomb) and the following words stand for “ introduction,
translation, and commentary, with an added life in brief of U.F.”.
In the text, the Muses become Xebbiet lehiema (fit. inspiring girls).
Foscolo s rather unhappy life until the end came at Turnham Green
is well told.
C. L. D.
M ARD TA Qari Malti. By P. P. Saydon and Jos. Aquilina. pp. xvi,
223. 8vo. Malta ; Empire Press, 1936.
This IS an anthology. The matter is mainly from the little magazines
-Mahi, Le%en il-Malti, a long-defunct periodical Is-Seb% (The Dawn),
and trom the Maltese equivalent of “ Old Moore’s Almanack ”. There
DELL TA’ L-IMGHODDI
1177
are also some extracts from books. In X'rajt u xi smajt (what I saw
and heard) Mgr. P. Galea tells of a visit to Coney Island in 1926,
where he saw the Witches' Cave (il-ghar tas-sBarijiet), the ponies
(ponijiet, i.e. the Engbsh word with the Semitic plural), boys ; lebsin
il-qalziet tal-ghawm, i.e. clad in swimming-slips, where qalziet is
the usual plural of the Italian calza, which, like many foreign words,
has quite changed its sense in becoming Maltese, and the wooden
race-horses : zwiemel ta‘ I-injam ghat-tigrija, where zwiemel is the
broken plural of ziemel, a horse, from primitive zamilah, modern
Spanish : acemila, a beast of burden, and 1-injam (wood) is from
Italian legname, with the initial treated as the article. The delights
of Luna Park were left behind '■ about 11 p.m. which is, in Maltese,
f 'xi 1-hdax ta’ bil-lejl. Here hdax is the Maltese reduction of primitive
al-hadi ‘ax(ar).
C. L. D.
Dell ta’ l-imohoddi. By A. V. Yassallo. pp. 39, 8vo. Malta:
Empire Press, 1936.
In Malta poesy grows apace. This is a collection of twenty-two
original pieces and an adaptation of one by T. Moore, all in various
metres. In the title, which means Shadow of the Bygone , dell
corresponds to the classic zillu, and imghoddi is the participle of the
Form II verb ghaclda " to pass ”. I once read somewhere that the
^laltese peasants, ev'cn now, sometimes indulge in rhyming bouts.
If that be true it might be a good thing to secure phonographic records,
for there is still much to be learnt from the lips of the bdiewi .
Some of the specimens of rustic prose collected by Hans Stumme had
an Urwiichsigkeit lacking in the more polished extracts, leading one
to wonder whether there are not many humble poets blushing unseen
in the countryside.
C. L. D.
S. Guann Bosco, Sajjed tal-eruieh. By A. M. Galea. From the
English by F. A. Forbes, pp. 176. Small 8vo. Malta : Salesian
Press, Sliema, 1936.
This is a life of St. John Bosco. The thoughtfulness of the future
saint for his Jewish comrades at school, which led him to pass on to
them the notes of the master's lessons on the Saturdays, when they
1178
REVIEWS OF BOOKS
were prevented from attending school, is a happy trait (p. 26). ■; -
complete list given at the end of the book of all Maltese Salesi.i -
past and present, strikes one as very exiguous. The fact is that >
Maltese as a whole do not take readily to missionary work. T' ■-
book, like all the many others of Comm. Galea, is printed in the spei '
type preferred by this writer, which, I believe, is used now only by lu;
c' L. D.
Taghlim Nisrani fuq il-Katekismu tal-Kardinal P. Gasp.\rk
By Dun Karm Farrugia. pp. 158, 8vo. Malta : Governnieii
Press. 1935. 4d.
The first edition of the Taghlim Nisrani, or, Christian Doctrin
in Maltese, came out in 1752 in the episcopate of the French Bishc.);
Paul Alferan de Bussan. In the British Museum there is only tb'
second edition published in 1768 in Rome (“ fisuk tal-Minerva " /
A third edition, with an enlarged title, was published in 1789 in Malta,
in the “ Palaz tas-sultan ”, the Grandmaster being styled Sultan,
The translator was a priest who spells his name in the second edition
Wzzinu. The Nihil obstat was given by a missionary named Ghebejer.
As explained on the title-page, the Maltese is based on the translation
made for the Maronites of Syria by order of Paul V who died in 1621.
In the British Museum there is a copy of the Doctrina Christiana in
Arabic and Latin for use of the Maronites, published in Paris with the
Imprimatur of the Master of the Sacred Palace in 1635, and comparison
puts it out of doubt that this was the work used by Wzzinu. The
Maltese follows the Maronite Catechism word for word, and, where
it departs from it, is only where western usage differs from the eastern.
Thus it omits to demand abstinence from wine in Lent, says nothing
under baptism of circumcision, and omits the doxology after the
Paternoster (though it had been in use in Malta not long before).
It agrees with the Maronite even in adding to the formula “ In the
name of the Father, Son, and Holy Ghost ” the words “ One God '
(Alla wiehed). It retains the construct case in such phrases as f’sieghet
mewtna (in the horn of our death), Xirket il-Qaddissin (the Communion
of Saints). But the period called the Great Fast (sawm d-kbir) it calls
Randan and also calls Easter Ghid il-kbir, two terms that still remain
in use.
In the course of the nineteenth century, new editions of the
Catechism were caUed for, and changes, chiefly by way of additions,
TAGHLIM NISRANI FUQ IL-KATEKISMU TAL-KAEDINAL P. GASPARRI 1179
were made. But, as the first work printed in Maltese, and one which,
by the nature of its contents, circulated widely, it was respected
as a “ classic For it must be remembered that for ages Maltese
was a purely spoken tongue, that it had nothing to do for it what the
Latin Vulgate did for the Romance Tongues, Luther's German Bible,
the Authorized Version, or the Quran for German, Enghsh, or Arabic,
or the various Divine Liturgies for the Greek, Syriac, Armenian,
Ethiopic, Coptic, or Slavonic.
The latest Catechism is a new book, based on the Vatican Catechis-
mus CathoUcus, yet usually keeping to the old Maltese wording. It is
rather a pity, that, in the eleventh article of the Creed, the old Maronite
text has been departed from ; inqumu minn bejn 1-imwiet . . . jilbsu
gisimhom . , . bil qawwa t'Alla (they will rise from the dead, put on
their bodies [as a dress] ... by God’s power).
Maltese, like Saxon English, is poor in abstract words. It has
words for apple, date, fig, grape, mulberry, but none for fruit, and
for ass, horse, dog, etc., but none for animal. Yet the Catechism has
Maltese names for all the seven “ Capital Sins For technical words
like grace, indulgence, purgatory, there were none available. I regret
that more use was not made of “ doublets The old Catechism had
“ tentati aw imgarbin for “ tempted and the new one has (p. 79)
■■ inkomplu jew intemmu ”, for “ perfected ”, but this should have
been carried much further. I was distressed to find that, on p. 94,
“ xbejjen ” the venerable old term for “godparents’ was not used.
Some of the new renderings are, however, quite good. The General
Judgment is Haqq ta‘ kullhadd, the Particular Judgment: Haqq
ta‘ kull wiehed.
The new Catechism, no doubt advisedly, omits a number of quaint
remarks found in the old, for instance, that, because the Creed was
composed by the twelve Apostles, therefore it has twelve articles, or
that the Sacrament of Extreme LTnction enables the soul to fight
against the devil at the moment of its separation from the body.
One oddity retained, even though it involves a departure from the
text of Holy Writ, is in the third Conmiandnient, which runs ; Thou
shalt observe the Sundays and the holy days (Hares il-hdud u 1-btajjel
(days, lit. “ empty ” of work) ).
On p. 73, breach of an abstinence-day is termed ” dirtymg oneself
with meat ”, the verb being nigges (II of olden nagisa). There is an
ancient Semitic tradition behind this use of the verb. On p. 20, the
second motion of the hand in the sign of the cross is said to be fuq
1180
REVIEWS OF BOOKS
is-sider, whereas the old Catechism, quite accurately, said tahr
sdirna, below our breast.
For Christian the Maltese use the word Xisrani (i.e. Xazarene).
In Maltese there still exists the term Grixti, which seems to be a nisbeh
formed from the name ‘ Christ I suspect that the word goes back
to the Middle Ages when the Christians were still a minority in a
Moslem community. Grixti now means much the same as rabli,
i.e. a rustic, retiring person. For its change of meaning compare the
Latin word paganus.
The new Catechism teaches the duty of forgi\dng offences. From
common report it seems that an exception is made against the unfaithful
wife, even when repentant, and, what is worse, that a man who is
disposed to overlook his wife’s folly tends to be scorned by his neigh-
bours. This is, however, an old complaint, for Megiser over three and
a half centimes ago noted how the Maltese women, when they were
allowed out at all, were compelled to veil themselves and to dress in
hideous garments all through the jealousy of their men-folk.
C. L. D.
Gabra ta Ward. L-ewwel sena (pp. 30), it-tieni sena (pp. 44j,
IT-TIENI KTIEB (pp. 98), IT-TIELET KTIEB (pp. 125), IR-RABA’
KTiEB (pp. 160). 5 vols. Large 16vo. Oxford University Press,
1935, 1936.
These are graded reading books for use in the Malta elementary
schools. The matter, in prose and verse, is taken from reputable
Maltese writers. Interspersed in the text are some Maltese proverbs.
Kelb rieqed la tqajmux, the native version of “ Let sleeping dogs he ”.
Xemx h ma ssahhanx ahjar il bard minnha (Better cold than a sun
that doesn’t warm). Bil-qtajra 1-qtajra tintela 1-^ajra (little strokes
fell great oaks), lit. tiny drop and tiny drop and the little jar is filled.
II kliem qabel tghidu iznu u qisu (before uttering a sentence weigh
and measure it).
C. L. O.
A Grammar of the Maltese Language with Chrestomathy and
Vocabulary. By Edmund Sutcliffe, S.J., Oxford University
Press, 1936. pp. xvi, 282. SmaU 8vo. 6s.
The Qcw Grammar is a most useful book, small but tightly packed
ur^hing, in highly idiomatic Enghsh, the solution of a whole
y of conundrums. Quite a number of items seem to me original
A GRAMMAE OF THE MALTESE LAXGUACJE
1181
contributions, at least I cannot recall having seen them before in
print. If some of the suggestions made fail to bring con\TCtion they
were well worth the risk of making. In the case of all languages.
Grammars and Lexicons made by foreigners are often more enhghtening
than those made by natives. The native is too much disposed to take
things for granted ; also it is hard for him to put himself outside
himself, a difficulty which in the nature of things does not exist for
the friendly outsider.
It was a good idea of the author to give a little up-to-date biblio-
graphy of Sicilian wordbooks, for just as a knowledge of Maltese
is a sine qua non for understanding Old Sicilian, so also is some
acquaintance with Old Sicilian and Old Spanish for the interpretation
of Maltese. The connection between the language of early medieval
Spain, Sicily, and Malta is a very close one, so much so that the three
dialects seem to have formed a group apart. It is also all to the good
that the author has not too narrowly confined himself to grammatical
considerations, but, has, in the case of many words, suggested
etymologies.
If I might single out anything for special praise it would be, apart
from the admirable treatment of the verbs, the manner at once simple
and illuminating in which the grammar deals with the accent or
stress. In the many words containing tlie digraph le (standing for
primitive a), like musbieh (lamp), halliel (thief) I note that he puts
(the accent on the i M'hereas Falzon puts it on the e ; the latest way of
marking the stress (on the i) comes nearest to the pronunciation.
By the way, in the matter of stress it is obvious that the laws of
accents must have changed in ilalta ; formerly it must have been
the rule that the antepenultimate took the stress, explaining how
barakah became barka, qasabah qasba. and waraqah werqa. Kov,
it is the penultimate that takes the stress, even in such a v'ord as
inbeda (began, p. 122).
The alphabet calls for a few remarks. Niggez (to prick, p. 3)
corresponds with form 2 of the old verb nakaza meaning the same ,
interchange of G and K is fairly common, as in Hebr. gamal, Greek
kamelos. Maltese gennen (shelter), is simply a variant of M. kennen.
In some words like gerfe.x (bungle), geddum (snout) the g-sound is
shared by Maltese with Jlaghrebinic. G. also sometimes stands for
Q as in words like gargar (roar), garr (coo), gemgem (mutter), gerrern
(gnaw). On p. 3 there is a mistake and the word gelgel (crack, split)
should be spelt with the soft g, and the suggested derivation must
1182
REVIEWS OF BOOKS
therefore be wrong, though it might be right for the word geli .1
(rush of water, p. 250). I recall for the sake of those who mav !ia ■
forgotten that, in Maltese, g stands for the Arabic Jim, x for Shi-
J for y, and g% for Ghain and "Ain. Under the letter k it might ha\
been pointed out that for some reason it seems to claim a following
as in the verbs kifes (eclipse), kines (sweep), kiseb (earn), kiser (break
kiteb (write), kixef (uncover). The grammar perhaps rather overdo-
the euphonic m . Wicc imb wicc (p. 254) is most likely for wi-:
magh (or, m ) wicc. The grammar felicitously instances (p. 28:'
some Spanish words like hombre where a gate-crashing b also make ■
its appearance after m. In the word mbagfiad the first letter is no
an emphatic ’ but simply an abbreviation of min(n).
In speaking of the letter Q it might have been noted that it-i
peculiar pronunciation in Malta has led to its being confused witl^
the Gh. Wisq (very) should by rights have been written wisgh
bis-serqa, when it means “ hastily ”, should have been written bis-
sergha, from the verb, now lost in Maltese, meaning “ to be quick " ;
qUla (fierceness) stands for ghUla, written with the rghain. Xieraq
(it is right and proper) were more correctly spelt xieragh (with the
Ain), as the word comes from the verb, lost in Maltese, xaragha
(to make a law).
Failure to find in the British Museum (p. 187, n.) the work of
Megiser that most enterprising researcher who visited Malta in 1588
was seemingly due to a German book having been looked for under
the Latin translation of its subtitle. The book was first published in
1606, again in 1610, and a third edition in the following vear. The
last edition is in the British Museum ; the title is Propugnaculum
Europae, Warhaffte . . , Beschreibung der Africanischen Insul Malta
. . . widerumb getruckt . . . Anno 1611. It is quite true that Megiser
gives as the Maltese word for the number “ nine ”, tesa, and, for
nineteen, tesatasch, but, then, he also gives as the Maltese for the
words hand, arm, house, and bed ; it, tria, tar, and sotto (now spelt
id, dnegh, dar, and sodda respectively), and writes the place-name
Uingli, (Hal) Tingli, all simply witnessing to the congenital German
diflSculty of distinguishing d from t. Why the Maltese should use
isgha for ten instead of the expected *tisgha is beyond explanation.
Usually the Maltese tend to use t for d, as in hotba, tafagh, tibek,
igiega (mstead of hodba, dafagS, dibek, and digiega, meaning
respectively : hump, push, grind, hen). Sicilian shared the same
>ng, and wrote ticchiena where Maltese writes dikkiena (bench).
A GRAMMAR OF THE MALTESE LAXGUAGE
1183
which is nearer to the spelling of this old Iranian word (dukkan).
Yet in the Maltese word daqs (for taqs), or, ghaddas (for ghattas ;
plunge) the change is exactly as in the word for “ ten
By the way, Megiser's little vocabulary, for all its misprints
has an interest of its own. For instance the words for two, three,
thirteen ” are fne, fliesan, flietash, which seem to show that the soft
th sound was stdl heard in Malta in the sixteenth centur}’, the sound
th being rendered by the letter /. For the word " gold " he gives veheb,
also tending to show that the initial soft d was still accurately
articulated. Yet again, for ” rich ” he gives rani, showing that the
initial rghain was still heard.
The new grammar (p. 196) suggests that the adverb of time il,
as in ilu (since he) ilek (since thou) is a reduction of hin (time) followed
by the preposition and personal suffix. This seems a fairly likely
supposition, though I cannot think of any other case where Maltese
has dropped the letter h.
On p. 55 the use of the word gmiel is well explained. Sbafit gmieli
(I awoke in fine fettle) has analogues in the chaste classical language.
On p. 61 and 245 the author is well justified in suggesting the Spanish
chico as the source of the Maltese clden (to grow small), where the
final n does not belong to the root. On p. 135 the Semitic origin of the
verb staUa (stop, rather recalling our English word to stall ”) is
vindicated. The verb, of which it is the X form, no longer exists in
Maltese. Ciera, another word anomalously written for *xiera, from
an original sarah (form), is also Semitic.
The use of lil to denote the objective case is ably explained (p. 169).
Elsewhere, the preposition lil usually stands for older li, like lil
kollhadd (p, 258, to everyone) for li-kull ahad. The relative li, illi
(p. 182) may stand, not only for ‘‘ who, which , but also for the
older particle an. Nifhem illi, in the old Catechism, corresponds to
afhamu an (I mean that . . .). A relic even remains in Maltese of the
old particle inna (verily), in expressions like kif inhu, how is he ?
(p. 181), daqsiex inhu (how huge it is !) (p. 198) reduced from taqs
ay xay inna-hu. It is rather unfortunate that neither the existing
lexicons nor the present grammar tell us anything about the
prepositional verbs. Yet the matter is important as the preposition
often determines the sense of the verb. Such prepositional \ erbs are.
e.g. naf bi (I know of), halla minn (let go), jisimgfiu minn (they obey),
sella ghal (he saluted), dahal minn (he entered). In the case of the
verb gieb (he brought) the preposition has been embodied in the verb
1184
REVIEWS OF BOOKS
exactly as it has been in Maghrebinic. It is to be hoped that '.e
promised new lexicon will make up for this defect in the exi.'ti' it
apparatus.
The grammar tells us that the genitive exponent ta' (or, ta i.
tal, etc.) is abbreviated from matagh, like Hal in place-names fri 'u
mahall (tribe). That is true, but it is also true that the t-marbaui
of the construct case of the many Maltese' nouns ending in aih) Las
come to play so important a part that it has acquired the meanm:
of “of”. In the old Catechism bi-qawwa t‘Alla, corresponds wita
bi-quwwati llahi, whilst ti^ s-saltna tieghek (thy kingdom conn
is the same as tigi saltnatek of the official version.
Another word, spelt ia, means “ give ”. On p. 139 we are tol'i
that it stands for ghata. That is correct sometimes, but not always.
Whenever it is a prepositional verb, as in ta biz-zewg (to kick, c}i.
Badger's Lexicon s.v. Kick), it stands for the old verb ata bi. On
p. 184 the author discusses phrases like taha ghax-xorb u ghal-loghob
(he took to drink and gambling). The prevalent native opinion is that
this stands for ta ha (ht. he gave it [his nih, or, soul]) to drink, etc.
But in some other phrases one wonders whether there is not here a
remnant of some verb, now lost in Maltese, hke taha, or taha.
There are several compound words. Jigifieri (that is to say, p. ‘-156)
used to be written as two words : jij^ jfieri, the latter word being
seemingly form iii of the olden verb fara (to show), and the measure
the same as in imieri of the form iii verb miera. Another composite
word is liema (p. 173, 182) which stands for h ma, and h ei ma. The
final ma is the neuter relative when the word is in the interrogative,
but sometimes the negative, as in liema bfialhom. Yet another com-
pound word is kallajamar, obviously for k-Alla jamar, whether the
initial k stands for the olden ka (as), or whether it is for jek (if). Jek
itself is identical with the old Quranic shortening of yakun. In Old
Spain yakun was used in the sense of “ if”. Pedro de Alcala wrote
ikin ente neseit (if thou didst forget).
In most languages we find foreign words being modified in form
or meaning by native influences. We have only to think of the English
trade-name (sparrow-)grass for asparagus. In Maltese the word
dmir (p. 247) (duty) is obviously the classical damir (conscience), yet,
in sense, it has been assimilated to the Itahan dovere. In the phrase
arma, ghid (come, tell 1) (p. 270) the word armi, imperative of rema
(launch out, throw) has been coloured by the Italian verb armare.
Kunjornok (thy surname, p. 258) combines a variant of “ cognomen ’’
A GRAMMAR OP THE MALTESE LANGUAGE
1185
with the olden word kunya. Missier (father) combines a Norman term
messire with the old word musawwir, the Quranic title for God which
perhaps approaches closest to our conception of ‘’father”. Mantar
(plural mnatar, p. 259), a frock, corresponds both with mantile and
with mamtar, plural, mamatir (a raincoat, from matara, to rain).
Some foreign words, though retaining their old meaning, have under-
gone an odd change. French assiette, has become as-siett, the first
syllable becoming the article.
There are also some interesting duplicates, where the same word
differently spelt has assumed different meanings. Minn jeddu
(spontaneously, p. 256) is clearly a variant of minn idu (off his own
bat). Diwi (an echo) is a variant of daghwa (which now in Maltese
means a blasphemy). The original meaning of the verb was to call,
invoke (God). OfH ghalik means bravo ! It is a reduction of ghafja
ghalik, on which see Dozy. Then we have kamla (a clothes moth,
p. 257) and qamla (a louse, p. 268). The reason why the former is
differently pronounced and written is that it is taken from the SiciUan,
but the Sicilian is itself from qamla which means not only a louse
but a cheese-mite, mealworm, or, grub of the clothes moth. In the
case of several words one Semitic one has been coloured by another
since lost in Maltese. Ghax irid (lit. because he wishes) is used in the
sense “ wilfully Here the existing Maltese verb seems to have been
coloured by the old verb radiya (be pleased). Errur, might have
been written gherur, and the verb, arra : gharr(a) ; the comparative
meaning “ worse ’ (agharr) is from the same verb. Jarralek il-hsieb
ta’ rasek (you will find out your mistake, ht. the thought in your head
will mislead vou). Here the Semitic word gharra (with the rghain)
has been affected by the Latin words : errare, error. Obda (to be
submissive) is simply the olden word ghabed coloured by the Latin
word obedire.
There are a few instances in which I disagree with the new grammar.
The author seeks to explain merhla (a flock) by bringing it into con-
nection with Arab, rihl (ewe-lamb), and implying that it is a noun of
multitude. That will not do, apart from the unlikelihood of a flock
made up of ewe-lambs, he has failed to notice that the old Maltese
lexicons spell the word with a h and not with a Merhla, plur.
mriehel, corresponds with Sicilian Arabic marhalah, plur. marahil.
It is true that the sense, from the context, is that of fold rather than
of flock, but the Maltese sense is also inherent in the root-verb. There
is a famous instance in the Fourth Gospel where the Vulgate has led
1186
KEVIEWS OP BOOKS
astray several of the Versions into writing “ one fold and o:i
shepherd”, where the Greek has “flock” and not “fold". The-,
again, idda (shone) is put by the author in the same case as itk
(leaned, p. 137). His reasoning is : Since itka is obviously form v
of the verb waka, why should not idda be form viii of the verb da
The proposal is plausible, and, formally, I see no objection to ir
But the fact is that, whereas the vih itka is in wide use outside of Malt.;
there is no trace of da having been used anywhere in form viii. Else
where it is form iv (ada) which expresses the Maltese sense of idda
whilst mudi stands for the corresponding Maltese adjective niiddi
(bright). It has also to be borne in mind that the Maltese word idda
also means “ to offer a gift ”, in which sense it stands for form ii of
the verb ada, a dialectic variant of the verb hada. Maltese makes no
difference between d and d. Tbe initial i in Maltese is accounted for
by the rule by which such measures as aqtalah, aqtilah, become in
Maltese iqtla, as in ilhna (voices) isqra (hawks), izmna (times), istla
(buckets).
On p. 61 the word ahjar is said to be in common use in the sense
of “ better ”, but it is denied that it is really a comparative. It is
given as the vulgar comparative in Badger’s Lexicon (s.v. better).
The lengthening of the final syllable can be matched in the word
gharghar (juniper-tree) where the final is also lengthened in Malta
and Old Spain, though it should be short.
I notice that, following Falzon, habba is rendered as “ iris of the
eye ”. Since the comparison is to the pip of the pomegranate it ought
rather to mean the pupil, which is the meaning as.signed to it by
Dozy. The more usual word for pnpil seems, however, to be mimmi
tal-ghajn, mimmi being the equivalent of the old word mimi (a baby).
It might have been as well to say something of the groups of
secondary verbs, for instance the maqtala(h) verbs some of which
Malta shares with Old Spain, where they also were usually found in
quadrihteral form ii. Tmannas (to grow tame) is from the same
root as stines, tmashan (be furious, p. 276) is from sahan (to warm),
tmaqdar, tmashar, tmattar are respectively from the verbs, lost in
Maltese, qadura (was imclean), sahira (laughed), and atara. Mekkah
(which should have been written mekhah, to dirty) is from kahh
(merd). Tmerzaq and immenzel seem of similar formation though
the verbs have acquired quite new meanings, and the same is true
of the piuely Maltese verb meclaq.
Then there is a group of denominative verbs, qatlat(a), made up
A GRAMMAR OP THE MALTESE LANGUAGE 1187
of nouns ending in the t-maibuta, which, in these verbs is hardened
into a ta. Such verbs are common in Maghrebinic and Egyptian.
Harbat (to ruin) is from herba (a ruin), qalfat (to calk), sarbat (to
put in a row) seem to come from nouns meaning “ tow ” and a “ file
of beasts '). Carrat (to tear) is a secondary form of carra, meaning
the same, on which see Mar^ais, Tanger. Finally there is the barbarous
but highly interesting measure qatlar, of verbs formed on the Spanish
model by adding a final ar to a noun. In Spanish we have many verbs
like aceitar (to oil, from zeit), azotar (to whip, from sawt), acequiar
(to drain) from saqiya. Corresponding to these, in Maltese, we have
bazwar (to rupture, from bazwa), izzakkar (to belly out, from zaqq).
It would have been as well to explain that foreign names of officers,
adjectives, etc., often take the nisbeh form with suffixed i giving them
an oddly plural look, thus, interessanti is a singular adjective, whilst
Kuntistabbli (p. 258) means a single policeman.
Some of the extracts in the Chrestomathy contain phonetically
written words which should have been explained in the vocabulary,
for instance, under tqieq (flour) a reference should have been made
to pp 1 and 53 where the word is accurately spelt with a d. On p. 253,
zgiega (glass) has been misspelt. On p. 17 the printer's imp got into
the text during the machining and badly bedevilled it. It is a pity
that the sections are not marked on each page, because the cross-
references to sections are very numerous and involve a lot of page-
turning and plenty proximate occasions for blasphemy.
The arrangement of the book is quite above reproach, and the
text, necessarily intricate, is most carefully done. One can understand
the author's enthusiasm for Maltese. It is the only dialect of its
family which has now for quite a long time been the object of intensive
study. There is also an especial reason why it should appeal to an
Englishman, for the Semitic and Latin play in it a part so like that
played by the Teutonic and the Gallic in Enghsh itself. Just as it is
to the Saxon that Enghsh owes its strength, and just as English hearts
warm to a speech or song in the measure that the Saxon element
predominates in it, so the Maltese in respect to the Semitic. The
native words (Saxon : ox, or Semitic : baqar) mostly have homely
pictures linked up with them, whereas the foreign word often conveys
nothing but a bald idea.
C. L. Des.soulavy.
Vi II,. vm. P.VET 4.
76
1188
REVIEWS OF BOOKS
\V kD.
bo
•th
he
he
>0
Ax IxTRODucnox TO THE Ibo Laxgoage. By Ida C.
pp. xiii, 215. Cambridge ; Heffer, 1936. 6^.
r. I. C. Ward’s new book throws a flood of light on th.
anguage. But its significance is greater than that. Together
her recent book on Efik it is a systematic attempt to Ivplor.
i^amihar temtory of tone and to elaborate a technique for
study of tius little understood linguistic element. As such it ca
tail to be of mterest to linguists generally. And to the student of
l-ITi Students of other tone langu .
° can be applied to any langu.;
^ ^ tonal has for long been as well known as
mphcatio^ of the fact have been little understood. This new anah i
shows in the first place not only that certain words are distinguish
semantically by their tone forms-for instance oie f '] “ male a
f ' rni, that each word in isolation has its own t(
orm. here are at least five tone classes of disyllabic nouns, two
monosyllabic verbs, and four of disyUabic verbs, and so on.
u as mth any linguistic element what the tones do is mojj
mportant than what they are and the central theme of the boo,:
is the way m which tones enter inseparably into the whole functionin;
instance such as m “ head ” an l
hange these tones m combination and become m - -] th ■
fmt/ ^ formations that th.
unjmrtant part played by tone emerges most clearly. There is on,
verb form which may be either a habitual tense, a present participle,
mmttT 1“ °‘her respects it
[rcZ r I ‘^e indicative and sub-
botrwT ! '' ^ “dicated
what SLss atdTe^iblv ^ ^
MnltThl wy>,aw. s„.n that 30-caa«i pthnitiv.
ignored. ’ *’*’ means of expression have probably been
A study of the book will reveal flio x ^
involved in the work. The method ^ of patient analysis
number of texts from r. y employed was to take down a large
to ocomate phonetie nomtion
AN INTRODUCTION TO THE IBO LANGUAGE
1189
both for sounds and tones and then to analyse the way in which the
tones worked in their various contexts. By comparing a large number
of usages it was possible gradually to formulate the main rules of tone
behaviour. The process, part of which the reviewer had the good
fortune to witness, was more enthraUing than any detective novel.
A point of particular interest is the statement that, so far as the
author’s investigations went, tone seemed to vary less than other
linguistic elements from one dialect to another. It is greatly to be
hoped that she may pursue her researches further into this field.
A number of interesting questions emerge from the book. For
example, do the tone-classes of nouns have any other, non-tonal basis
of classification as in Bantu or Fulani ? Is it, for instance, just chance
that isi [~ "] “ head ” has high tones whereas ala [- -] “ ground ”
has low ones ?
The whole question of the function of tone in Ibo needs further
consideration. And one may perhaps hope that if the author of this
book extends her investigation to yet other West African languages
she may throw light on the general function and behaviour of tone in
language, if such there be.
A reading of the chapter headings of this book reveals the welcome
fact that certain forms of the verb are simply labelled the nu-form
or the ro-form instead of being pressed into ready-made grammatical
categories. The use of categories imphes an excursion into semantics
which can only be justified by ample evidence. And the Ibo verb
needs much more study than it has at present received before its forms
can be classified. In many ways it tends to emphasize aspect rather
than tense and the exact significance of the various forms needs a
great deal of careful contextual investigation, as the author has
realized. Her book had to be planned for practical as well as scientific
purposes and a definitive handling of many situations had therefore
to be postponed.
The system of tone notation used in this book is of practical
interest to all field workers whether linguistic or anthropological.
Of the various systems adopted up to date it seems the most efficient
and least unsightly. And the whole method of the book shows, if any
showing is necessary, how vital is the recording of texts in an adequate
script both for sounds and for tone. How far tone notation should be
used in the printing of books is a different matter. In such a thing as
the translation of the Bible for instance, passages that are particularly
obscure because of alternative tonal possibilities might have a short
1190
REVIEWS OP BOOKS
elucidating tone pattern in brackets in place of the present inadequ ite
low or high tone symbol. But the question is not altogether simple and
needs careful consideration. No one would presumably suggest t’ at
books in the vernacular should adopt fuU tone notation. The native
reader does not need any such cumbersome device. But the presc-i.t
reviewer would without hesitation recommend that any book or tevo
published for the use of students of the lang^uage should adopt tone
notation in fuU. Personal experience has shown that toneless tex*'
are more or less useless except to an advanced student of the languaei
The present book with its te.xts with full notation is invaluable.
The publishers of the book are to be congratulated on then
excellent production.
-Uargabet j\I. Greex.
Baxtu Lixguistic Ter.min’ology. By Cle.ment M. Doke. pp. '237
London : Longmans, Green, and Co., 1935. 6s.
The Introduction . — In his introductory section on Bantu Linguistic
Terminology, Professor Doke has raised a question of great importance-
concerning principles which underlie the correct composition of a word.
The need for some measure of uniformity in writing a Bantu language
is recognized by everyone interested in linguistics, and on p. 5 the
author expresses the hope “ that interest and seriousness will be
stimulated and some degree of uniformity aimed at as a result ' ■
The hope referring to interest will be abundantly fulfilled, for his
book will bring about much discussion, and in the course of discussion
much interesting matter concerm’ng Bantu linguistics will emerge.
The result will undoubtedly be to aim at “ some degree of uniformity .
but it remains to be seen whether the conclusions arrived at will be
acceptable to all.
In dealing with the composition of a Bantu word, Professor Doke
will find few to disagree with him that Bantu writing should be con-
junctive rather than disjunctive ; but whether conjunctive writing
should be carried further for the sake of uniformity is a question
needing detailed examination. For instance, Professor Doke points
out that -a in Swahili, whether appearing as na or with the concordial
prefix, is at present written conjunctively when associated with a
pronoun but disjunctively when followed by a noun. But is not this
as It should be ? -a and na are written disjunctively when followed
nouns, and conjunctively if the particle
o ows IS monosyllabic— generally pronominal in character.
BANTU LINGUISTIC TERMINOLOGY
1191
e.g. Pronominal : Alikwenda nacJio. Watu wake.
Nominal : Alikwenda na kikapu. Watu wa Hamiai.
If the emphasis is on the absolute pronoun, the full form is retained
and the na is then written disjunctively : Na yepe pia, for this is in
accordance with what Professor Doke himself savs on p. 17 : ‘‘ Stress
is the word-builder in Bantu.”
Turning to the next heading in the Introduction wc read on p. 24 ;
Taking the word as the basis of Bantu grammatical classification,
each complete word constitutes some ‘ part of speech ’ according to
its syntactical force. . . . There will thus be found six fundamental
parts of speech.” This statement is too highly controversial and
complex to be dealt with in a short review. It must suffice to say that
Professor Doke has yet to convince some of his readers why “ -a"
must he reckoned as a formative rather than as a word to indicate
prepositional relationship.
The subdivisions of the fundamental Parts of Speech are in the
main helpful. It is wise to let “ Adjective ” give place to the wider
term “ Qualificative ” and become a subdivision under it, for many
words in Bantu “ qualify ” substantives, but cannot be called
adjectives., e.g. Rifiki yangu sana (My great friend). Possessives,
however, should also appear as a subdivision of pronouns as well as
a subdivision of quaUficatives : similarly demonstratives should
appear in both categories.
e.g. Lete vyuo vyangu (Bring my books). Lete lyangu (Bring mine).
Chuo kick (This book). Hicho sikitaki (I don't want this one).
It is to be regretted that the neutral nature of many stems has not
received more attention when dealing with the question of classification
of the parts of speech, although there is a reference to it on p. 26.
The case for replacing “ Adverbs” by “ Descriptives ” is not
particularly convincing. The point in its favour is that it would then
include ideophones, for ideop hones do describe verbs. But the majority
of adverbs, other than adverbs of manner, modify the statement
rather than describe, e.g. Alifanya kwa ujinga. (He did it through
ignorance). “ Ktva ujinga ” does not describe the action.
In respect to the next heading “ The Rendering of Linguistic
Terminology in Bantu ” it has long been felt that there is a crpng
need for the introduction of a terminology in place of the conglomera-
tion of terms at present in use. The author sums up the whole situation
when he says, p. 4 : “ Some Bantu grammar written in a European
language is taken and terms translated or transhterated (according
1192
REVIEWS OF BOOKS
to individual preference) therefrom without any real inquiry into t
basic meaning or function of the terms.” And on p. 2 ; “ We nu
be prepared to blaze a new trail, use new terminology where necessary
And again on p. 32 ; “A great advantage will result, if at least t
main parts of speech have similar terms in various Bantu language'
The Dictionary . — Our thanks are due to the compiler for th
great contribution to Bantu linguistics. It supplies a long felt wai
both to the ordinary student and to the research worker, and w i
be of immense value for many years to come. It is impossible t'
mention more than a few among the very many terms and explanation'
which are particularly helpful. Chief among them are: absoluti
adjunct, apposition, concord, concurrent, conditional, definite
derivative, emphatic, enclitic, ideophone, implication, mood, partici
pial, potential, progressive, qualificative, subjunctive, syntax.
A right understanding of these and other terms is of primary
importance to a beginner, for they should do away with the confusion
which often arises in his mind as study progresses, due to the multipli-
city of terms used by various writers to explain one and the same thing.
Lack of space forbids a discussion of any of the controversial
points, but some of the explanations given in the dictionary will
need modification. Professor Doke is an eminent scholar and one
great enough to be ready to give way to “ surer descriptions ” (p. 36)
in the future as he has done in the past. For, as he says, p. 36 : ‘‘ Only
the very best is good enough for the languages and literature of the
Bantu.”
E. 0. A.
Kamusi ya Kiswahili yaani Kitabu cha Maneno ya Kiswahili
KIMETUNGWA. Na F. JoHNSON. Swahili-Swahili Dictionary,
pp. xvii + 261. London : The Sheldon Press, 1935. 2s. Gd.
The compiler is to be congratulated on bringing out this dictionary
which should do much to facilitate the study of Swahili, both from
an academic and colloquial standpoint.
To the European it provides an insight into the wider meaning
of words, and to the African whose mother tongue is not Swahili it
supplies a very urgent need.
The book should have a ready sale and prove the truth of an old
African saw “ Chema hujiuza ” (A good thing sells itself) — especially
at the modest sum of half a crown !
E. 0. Ashton.
TEN AFRICANS
1193
Ten Africans. Edited by Margery Perham. pp. 356. London :
Faber and Faber, 1936. 15s.
That Ten Africans bas been reviewed in a prominent position in
newspapers like The Times, the Observer, the Daily Telegraph, as
well as in journals dealing specifically with Africa, is indicative of
the wide interest which African affairs are attracting to-day and is
no less a tribute to the reputation of the editor. Miss Margery Perham.
The book consists of ten stories each the sketch of the life of an African.
The subjects of the stories are as widely separated geographically as
Nigeria and the Transkei, and in education range from men removed
from brutal savagery by one short generation only to the finished
product of an English school and an Oxford women's college. Each
teUs his own story : four of the ten write their history in English and
the others have told it in their own languages, and a number of well-
known workers in the African field have recorded and translated
these. Miss Perham has collected and chosen these representative
contributions and has written a short introduction to them. Her
collaborators have let the African talk and their talk rings true :
there has been no alteration of matter or style : from internal evidence,
it is easy to see that the original texts have been faithfully adhered
to, that sometimes the very idiom of the language has remained.
The writers for the most part are of that generation which has
seen the change from the old to the new regime and which has felt
the impact of Europe upon Africa and they frankly express their
opinions upon these changes. It is only fair to say that they are
ready to acknowledge the good as well as to deplore what they consider
the evils of European domination in Africa.
It is very much to be hoped that the favourable reviews which
have appeared will attract not only the speciahst reader but that
section of the general pubUc who would not be likely to read the
works of the social anthropologist and the administrator but who
form their judgments of African matters upon popular and often
somewhat slight books of travel. These stories give a living picture
of the life of Africa to-day and yesterday and should do much to
break down facile and often false generalizations about Africa which
are responsible for the popular attitude to these backward peoples.
The book should also arouse a sympathetic understanding for the
difficulties of those who have the administration of our colonies and
dependencies in their hands and especially for the District Officer
1194
REVIEWS OF BOOKS
who has to deal with the problems on the spot. One of the story-t- rs
reahzes that the D.O. may be “ worried by havdng so much to dt:
Africa has httle recorded history and still less written hy !ie
African himself. Here is a book which will prove a valuable contribu i jh
to the social history of an important period in the developmeir oi
the continent. As such, not only should the European benefit l)v le
reading of this book, but it should be made available to the Ain*
himself. It is to be hoped that the six stories recorded in the nm > e
languages may see hght in their original form. But all the stm. s
could very well be translated into a number of the important languav '
of Africa and thus form a welcome addition to the growing, but 4
small, body of vernacular reading matter.
I. C. Ward.
NOTES AND QUERIES
OM MAXI PADME HUJI
With the possible exception of the Muslim Bismillah, no formula of
Oriental religion is so common as the phrase mi tnani padme hum,
which appears countless millions of times in Tibet, engraved or
painted on rocks, stones, prayer-wheels, and tablets and written or
printed on paper. But whilst these words are wearisomely familiar,
there is no general agreement as to their meaning. It is commonly
taken for granted that rmni padme stand for the Sanskrit words
manih padme, “the jewel in the lotus,’’ and this is possible. But the
sense of such a combination is not clear. It might designate the Buddha
or a Bodhisattva seated upon a lotus-throne, or a Buddha born in a
lotus. ^ But if such be the meaning it is very hard to imagine how so
indefinite an allusion could have won the enormous popularity which
it enjoys in Tibet. A better explanation is indicated in her trans-
lation of the story of Dri-med-kun-ldan (Ti-me-kun-dan , in the Wisdom
of the East Series) by Mrs. M. H. Morrison, who writes (p. 13) : The
syllables Mani-padme represent in all probability the name of the sakti
of Chan-re-si.” ^ She gives no authority for this opinion ; but a
Tibetan charm which I examined many years ago seems to supply the
necessary proof, and I therefore print the text of it, with all its blunders.
Nama samantabhudadhani,® sarvabhavasantarani huni phat
pataya svaha, om, om bhu,® om padme, om padme hri,® om
padmabhudze, om padmalokite, om padmadzvala ‘ hum ghrig, om
amoghamanipadme, orn padmalotsana * hmu hum, om padmo-
snlsabhimale ® hum phat, om ah hurn.
1 For the orthodox explanation see page 1 of Dri-med-kun-ldcn's Xamthar in
En-glish, translated by K. Sumdhon Paul, in the Journal of the Department of Letters
of Calcutta University, vol. 2.i.
2 Chan-re-si is the phonetic spelling cf Spyan-ras-gzigs, i.e. Avalokitesvara.
® Probably to be corrected to Xamah samantabhadrani, on the analogy of bhamni,
rudrofii, etc.
* Read -santarani.
^ Read bhu,
* Read hrl.
’ Read padmadzvaJe.
* Read padmalotsane.
* Read padmosjusaviinale.
1196
NOTES AND QUERIES
Here the ^akti or female Power of Avalokitesvara is invoked r
t e destruction of malign forces in a series of epithets of which m '
contain the word padina, “ lotus.” The first epithet suffices to sii v
he connection with Avalokitesvara^ to whom there is a furr: >
CTenee in the epithet amoghamanipadim, for Manipadma is also < >
fF ^ The final and decisive proof of the connection
a or e y the fact that the formula om mani padme hum or mdah-:
IS commonly considered to be attached to the cult of Avalokitesvai
the Sadaksari Mahavidya, is represented as sitti: .
beside him when he is worshipped as Sadaksari-Lokesvara.* Thus t!
rmu a, originally addressed to Avalokitesvara's female Powei
as been applied generally to his cult, with which it has gradualk
spread over the whole religious life of Tibet.
epithet tnanipadmd, “ jeweMotus,” may be explamed from tlu
nnection. Sadaksari-Lokesvara and Manipadma are usually figure^ ■
^t our arms, two of them holding a lotus and rosary and the othei
^ointhe ahjali position, with a jewel in them ; sometimes the formci
o s in his hands a jewel, book, and lotus,^ and sometimes the spirit
.a. a sari also holds a jewel. Sadaksari-Lokesvara is commonly
attended by a familiar named Manidhara, “ jewel-bearer.” Thus the
lotus and jewel are characteristic emblems in this cult. Now one of
he epithets in our charm is amoghamanipadme, ” having infallible
jewel and lotus," a regular bahuvrihi compound ; and it is quite
P a ® from this or similar compounds the irregular TOunipud/Md
as etached as an epithet of the ^akti, who was now addressed with it
e abridged formula mn manipadme hum, while it was also in the
masculine form applied to the god.
B,
A NOTE OX KASSITE PHONOLOGY =
There seems to be no reason to doubt the usual « explanation
e ements hurna-, -buryas, and -bugas, which are found in
^ B. Bhattaeharyya, p. 183.
’ SadhanamaU, p. 33.
* iladhanamald, p 35.
Babylo^far iolo^. for advice on various points of
»W references ‘o IScrUt^S^' wiU^bffol^i 142 ff..
NOTES AiJD QUERIES
1197
Kassite proper names, as Indo-European. No adequate explanation
of the u in the steins of these three elements has, however, hitherto
been given. ^
The relevant data ^ are as follows : — •
(1) hurna- “ protege ” : bur - na - bur -ya-a's = ki - din - bel - mdtati
“ protege of the Lord of the Lands ” (King-List, No. 28) — Goth.
Olcel. OHG. OS. barn, OE. beam “ child " ; Albanian burs man ”,
OIr. barn “ judge ; Lith. bernas “ youth, boy ", Lett, berns “ child
(2) -Buryas = Hadad-Eimmon (storm-god) : u-larn-bur-ya-as = U-
dan-bel-mdtati “ child of the Lord of the Lands " (i.e. Hadad-Rimmon)
(King-List, No. 25) ; Burna-buryas (above) ; further, in the list of gods
on the British Museum tablet K. 2100 Rimmon is equated to “ Buryas
of the Kassites ” ® — Gk. Bopeas, OBulg. burja " tempest ".®
(3) -Bugas god-name : in Nazi-bugas (Pinches, p. 108, No. 21)
clearly meaning [my] protection [is] Bugas ” (cf. Vocabulary 38,
na-zi = sil-lu”' “ protection ”) — Skt. bhagah apportioner. Lord, name
of Savitar and of another Aditya ” Av. baya-, Sogdian ^ay, Middle
Parthian bay. Middle Persian bay apportioner, god ”, OBulg. boyh
“ God”, Phrygian Bayalos' Zevs 0pvyio^d
On the prevailing theory that the Indo-European element in Kassite
is Indian or Indo-Iranian. the u in burna-, -Buryas and -Bugas cannot
be explained. Ind. or Ind.-Ir. a could only appear as d in Kassite,
and there is no reason to assume that it would be transliterated other-
wise than with d in cuneiform.
I wish to suggest that this Kassite u represents not d, but 6 ; in
fact that Kassite burna-, -bur-, -bugas represent forms *bhdrno-,
^ J. Scheftelowitz, Zeitschrifl fiir vergleichende Sprachforschung, xxxviii, 261,
suggests that the « is due to the influence of the preceding labial, but this view has
not found general acceptance.
* The Kassite material is presented in a convenient form by T. G. Pinches, Journal
of the Royal Asiatic Society, 1917, 101-114 (cf. particularly the “ King-List ”, pp. 106-7,
and the “ Vocabulary ”, pp. 102-5) ; see also F. Delitzsch, Die Sprache der Kossder.
® A. Walde and J. Pokomy, Vergleichendes Worterbuch der indogermanischen
Spracheu, ii, 155 ff.
* C. Bezold, Proceedings of the Society of Biblical Archseology, ix, 377.
* The form ubryas, which is apparently equivalent to buryas (cf. Vocabulary 6
ub-ri-ia-as = UuAddu, Rammdnu ” Hadad, Rimmon ”), is obscure. For such a
“ metathesis ” — whatever its explanation — we may perhaps compare Vg-ba-ru
(? = normal Chi-ba-ru) ; see S. Smith, Babylonian historical texts relating to the capture
and downfall of Babylon, pp. 121-2.
* See J. Charpentier, I ndogermanische Forschungen, xxix, 378-9. The exact relation
of the Kassite and Greek sufiixes is not clear.
’ Walde-Pokomy, op. cit., ii, 128.
1198
NOTES AND QUERIES
h{h)or-, bhoffos. d is lactinp in cuneiform and it would tht ri ip
have been transliterated with li. As parallels it will suffice to r it. r.
cuneiform transcriptions of Greek such as An-ti-gu-nu-H,-,n ,
Pi'la-a-gu-ra{-a) ^cXdyopas—see K. L. Talkjvist. -Ass. m
Personal Names ' (Ada Socieialis Sdentiarum Fentuar, xliii. .
pp. 1.36, 181a , An-ti- i-u-hu-sH Avrloxos see. further. .M. Pur' •,
Babyloniaca, xv, 67.
If this view of the Kassite u be accepted we can hardly re^^anl ■
assite forms as originating in any form of Indo-Iranian in the ordit. .
sen^ of the word. For one of the chief characteristics of Indo-Irai.. -ri
IS t e change of IndE. o to Ind.-Ir. n. We must therefore assume ti it
ese assite forms originated in (a) Indo-Iranian of so early a d,.,.'
that the change of IndE. 6 to Ind-Ir. d had not yet taken pla.' ^
) m Primitive Indo-European itself or (c) in some other form "t
Indo-European which preserves IndE. 6.
Alan S. C. Kos>.
On 6 or bh see C'harpentier, loc. cit.
"n determine the date at which this Ind..Ir. change t." I.
£ <*- “ -'•rt'iinly took place after ft.
(J- tVackemagcl. '
with tK ^ that this latter change was contempordi.^
It discussion, this can afford us no more than a relative .Uft
lwhnfr.ll * * A Po^'tion maintained bv H. Jacob-nli'
1^. ^ 4 r/€r un/i f grofinnen. the lexicographical corrcsponrinu ' '
Arviin ” **^°^*^*'^ inno-Ugrian and Indo-European, often regarded as due to “ Pro
j j rimitive Finno-Lgrian cannot afford us anv assistance u
determining the date of the change of IndE. o to Ind.-lr. «. .See B. Collinder. ]»'!' ■
uralische^ Sprachgut and A. .S. C. Ro,,s, BSOS., viii. I, pp. 2i>7-234. 193.5.
BOOKS RECEIVED FOE REVIEW
Agixsky, Ethel G. A grammar of the Mende language. Linguistic
Society of America, Language Dissertations. Ko. 20. University
of Pennsylvania, Philadelphia, 1935.
Atulaxanda Chakrabaeti. Cultural Fellowship in India. Calcutta ;
Thacker Spink, 1934. Qs.
Bal Gaxgadhar Tilak. Om-tat-sat Sriinad Bhagavadgita Rahasya
or Karma-yoga sastra. Translated by Bhalchandra Sitaram
Sukthankar. Vol. i, Poona, 1935. Rs. 6.
Bixa’ox, Laurence. Chinese Art and Buddhism. Proceedings of rhe
British Academy. London : Humphrey Jlilford. 1936. 2*’. 6d.
Brajendranath Se.^l. The Quest Eternal. Oxford Lmiversity Press,
1936. 6s.
The Buddhacarita or Acts of the Buddha. Sanskrit te.xt edited
and translated with introduction and notes by E. H. Johnston.
Two volumes. Panjab L^niversity Oriental Publications, Nos. 31
and 32. Calcutta, 1935, 1936.
Caiger, Stephen L. Bible and Spade. An introduction to Biblical
Archaeology. Oxford University Press, 1936. Gs.
Chafik T. Chebata. Essai d'une theorie generale de Tobligation en
droit musulman. Tome premier ; Les sujets de I'obligation.
Le Caire, 1936.
CoBBOLD, Lady Evelyn. Kenya, the land of illusion. London :
John Murray, 1935. 10.9. 6rf.
Davids, Mrs. C. A. F. Rhys. The Birth of Indian Psychology and
its development in Buddhism. London ; Luzac and Co., 1936. 5s.
Delachaux, Th.. and Ch.-E. Thieb.aud. Land und Volker von
Angola. Verlagsanstalt Viktor Attinger, Neuenburg, 1935.
Escarra, Je.xn. Le droit chinois. Peiping, Henri Vetch, 1936.
Frs. 95.
Faddegon, Barend. Studies on Panini's Grammar. Publications of
the Royal Academy of Sciences, Amsterdam, Phil. Section, xx.x\dii.
No. 1, 1936. 75. 6d.
Fawcett, Sir Charles. The English Factories in India. Vol. i (New
Series), The Western Presidency, 1670-1677. Oxford : Clarendon
Press, 1936. 18s.
Grousset, Rene. Histoire des Croisades et du Royaume Franc de
Jerusalem. Ill, La Monarchie Musulmane et I'Anarchie Franque,
Paris : Librairie Plon, 1936,
1200
BOOKS RECEIVED FOR REVIEW
Haarhoff, T. J. Afrikaans, its origin and development. Oxf 1 :
Clarendon Press, 1936. 2s. 6d.
Harivam^apurana. Beitrag zur Kenntnis des Apabhramsa un i r
Universalgeschichte der Jainas. Herausgegeben von Lu ji;
Alsdorf. Hamburg: Friedrichsen, de Gruvter and Co.. 1 ’
RM25.
Htllelson, S. Sudan Arabic Texts. Cambridge University Pi .
1935. 12s. 6d.
Hla-Dorge, Gilberte. Une Poetesse japonaise au XVIIP sib
Kaga no Tchiyo-Jo. Paris : G.-P. Maisonneuve, 1936.
Ibn Hasan. The Central Structure of the Mughal Empire and -
practical working up to the year 1657. Oxford University Pie
1936. 18s.
JouRDAN, P. Notes grammaticales et vocabulaire de la langue Da> .
London : Kegan Paul, 1935. 6s.
Kalidasa. The Meghaduta of Kalidasa. Translated from the Sanski
by G. H. Rooke. Introductory poem by Sir Rabindranat
Tagore. Oxford University Press, 1935. 10s. 6d.
Kowalski, Tadeusz. Les Turcs et la langue turque de la Bulgarn
du Nord-Est. Polska Akademja Umiejptnosci. Memoires de k-
Commission Orientaliste. No. 16. Krakow, 1933. Zl. 1.50.
Kshitimohan Sen. Medieval Mysticism of India. London ; Luzar
and Co., 1936. 10s.
Labat, Rene. Le Poeme babylonien de la Creation. Paris ; Adrien-
Maisormeuve, 1935. Frs. 45.
Levis, John Hazedel. Foundations of Chinese Musical Art. Peiping:
Henri Vetch, 1936. 21s.
Luke, Sir Harry. The Making of Modem Turkey ; from Byzantium
to Angora. London : Macmillan and Co., Ltd., 1936. 10s. 6J.
Macarius. The Travels of Macarius. Translated into English and
printed . . . 1836. Selected and arranged by Lady Laura Ridding.
London : Oxford University Press, 1936. 6s.
Maratha History. English Records of Maratha History. Poona
Residency Correspondence. Volume 1 ; Mahadji Sindhia and
North Indian Affairs, 1785-1794. Edited by Sir Jadunath
Sarkar. Volume 2 ; Poona Affairs, 1786-1797 (Malet’s Embassy).
Edited by Govind Sakharam Sardesai. Bombay : 1936. 9s.
and 13s. 6d.
Mattodaira, N. Les Fetes saisonmeres au Japon (Province de Mikawa).
Etude descriptive et sociologique. Paris : G.-P. Maisonneuve.
1936. Frs. 25.
BOOKS KECEIVED FOR REVIEW
1201
Meinhof, Carl. Die Entstehung flektierender Sprachen. Eine
Untersuching. Berlin : Dietrich Reimer, 1936. RM. 6.
Mendis, G. C. The Early History of Ceylon. Second edition. Calcutta :
Y.M.C.A. Publishing House, 1936. Rs. 2.8.
The Mirror of Gesture, being the Abhinaya Darpana of
Nandikesvara. Translated into Enghsh by Ananda K.
Coomaraswamy and Duggirala Gopalakrishnayya. Second
edition. New York ; E. Weyhe, 1936. $5.00.
Mus, Paul. LTnde vue de TEst : cultes indiens et indigenes au
Champa. Hanoi : Bcole Fran^aise de I’Extreme Orient, 1934.
Patwardhan, Rev. S. B. First Gondi Manual. London : S.P.C.K.,
1935. 2s. U.
Radhakrishnan, S. The AVorld’s Unborn Soul. An inaugural lecture
delivered before the University of Oxford. Oxford : Clarendon
Press, 1936. 2s.
Rozwadowski, Jan M. Worterbuch des Zigeunerdialektes von
Zakopane. Polska Akademja Umifjetnosci. Memoires de la
Co mm ission Orientaliste. No. 21. Krakow, 1936. Zl. 5.
Selmer, Ernst W. Experimentelle Beitrage zur Zulu Phonetik.
Oslo : I Kommisjon Hos Jacob Dybwad, 1933. Kr. 3.50.
Sk0LD, Hannes. Materialien zu den iranischen Pamirsprachen.
Lund, C. W. K. Gleerup. London : Oxford University Press,
1936. 25s.
Spies, Otto. An Arab Account of India in the fourteenth centurjL
Bonner Orientahstische Studien, Heft 14. Stuttgart :
Kohlhammer, 1936. RM. 3.
Stopa, Roman. Teksty Hotentockie. Polska Akademja Umi^tno^ci.
Memoires de la Commission Orientaliste No. 26. Krakow, 1936.
The Sutra of the Lord of Healing. Buddhist Scriptures Series
No. 1. Peiping, 1936. 2s.
Thompson, R. Campbell. A Dictionary of Assyrian Chemistry and
Geology. Oxford : Clarendon Press, 1935. 21s.
Uttley, Freda. Japan’s Feet of Clay. London : Faber and Faber,
1936. 15s.
Wangelin, Helmut. Das arabische A'^olksbuch vom Konig azZahir
Baibars. Bonner Orientahstische Studien, Heft 17. Stuttgart :
W. Kohlhammer, 1936. RM. 15.
Zajaczkowski, a. Ktudes sur la langue Vieille-Osmanhe. 1. Morceaux
choisies de la traduction turque anatolienne de Calila et Dimna.
Polska Akademja Umifjetno&i. Memoires de la Commission
Orientahste, No. 17. Krakow, 1934. Zl. 6.
Kegan Paul, Trench, Trubner St Co., Ltd.
ORIENTAL AND AFRICAN BOOKSELLERS AND PUBLISHERS
CLASSIFIED STOCK OF NEW AND SECOND-
HAND BOOKS ON THE LANGUAGES AND
LITERATURES, ARTS. HISTORY, GEOGRAPHY,
AND ANTHROPOLOGY OF ASIA AND AFRICA
INSPECTION INVITED
FOR THE MOST COMPLETE RECORD OF NEW LITERATURE
on tKe above subjects
KEGAN PAUL’S LIST OF NEW BOOKS
Catalogue XXIX. China and the Far East,
now in preparation. Will be sent free on request.
Libraries and collections of books and journals purchased
38 GREAT RUSSELL STREET, LONDON, W.C. I
OPPOSITE THE BRITISH MUSEUM
— BLACKWELL'S
will be pleased to send their CATALOGUES of
ORIENTALIA
No. 377. SEMITICA, including EGYPTOLOGY, ASSYRIOLOGY.
ASIA MINOR, etc., etc.
No. 383. The History, Geography, and Languages of BRITISH
INDIA, BURMA, CEYLON, etc., including the SAN-
SKRIT and PALI portions of the Library of the late
Dr. JAMES MORISON, Librarian of the Indian Institute,
Oxford.
No. 387. The Geography, History, Topography and Languages
of ARABIA, PERSIA, TURKEY, and THE FAR EAST.
— — GRATIS ON APPLICATION — —
L L ’ S
Booksellers
OXFORD
B L A C K W E
New and Second-hand, Universit
broad STREET
INDEX TO VOLUME VIII
A
Achin, language of, 1 89 et seq.
Ahmet-Zeki Valtdi, Ein turkisches Werk
von Haydar-Mirza Dughlat, 985-9.
Ali Nihat. Dr., SeyM divanini ietril:.
Part 1, 259. '
Alphabets and phonology in India and
Burma, 517-546.
Alsdorf, L., The Vasudevahindi, a
specimen of archaic Jaiua-MaharastrT,
319-333.
Altindische und mittelindischc Jliszelleii,
823-834.
Aquilina, Jos., and P. P. Saydon, tVar/l
to,' Qari Main, 1176.
Arabic, history', 143 et seq.
Arabic theatre in Egypt, I, II, 173-187 ;
III, 99I-I012.
Arberry, A. J. (translated by), The
Doctrine of the Sufis, 1169.
Armenia, history. 61,3 et seq.
AsHTO^^ E. 0,. The e and o of LuGanda
and the o of Swahili. 1121-3; The
structure of a Bantu language with
special reference to Swahili. 1111-1120.
Ashton, Leigh (edited bv), Vhine.w Ait,
1134.
Ayesta. 405 et seq.
Awromani texts, 467 et seq.
B
Bailey, H. W., Hvatanica, 923-936 ;
Ir.inmn studies, V, 117-142 ; Ttaugara,
883-921 ; Yazdi, 335-361 ; (with E. H.
Johnston), A fragment of the Uttnrn-
tiiiitra in Sanskrit, 77-89.
Bailey, T. Gr.aiia.me, pholii-Marurii
Duha, 975-9 ; Poes Khari Boll mean
nothing more than Rustic iSpeceh ?,
363-371.
Banarsi Das .lain, .4 Phonology of Panjabi
ns spoken about Ludhiana and a
Ludhiani Phonetic Reader, 220.
Bantu language, structure of a, mi-
ll 20.
Bantu languages, 1121 et seq.
BAE.AYSiKoy. .4., Modern literary Hindi,
373-390.
B.arbofr, XEvn.L, The Arabic theatre
in Egypt. I, II, 173-187 ; III, 991-
1012.
Bare, K., Remarks on the Pahlavi
ligatures and 391-403.
Batakrishna Ghosh, Les formations
nominahs et rerhales en p dii Sanskrit,
214.
Beaujard. Andre (translated by), Les
Notes de Cheret de Sei Shdnagoid, 275 ;
Sei Shbnngoii, son temps el son oeuvre,
275.
Bengali, drama, 113 et seq.
Benvesiste, E., Les in.finitifs avestiques,
11,>5; ,Sur quelqne.s dvandva.s
.avestiques. 40.5-9,
Bhagavata Parana and the Ivaiikas of
G.andapada, 107-111.
Bhiinaha, 713 et seq.
Bird-names in the Indian dialects, 599-
601.
Blaehere, R., Abou T-Tayyib .41-
Molannabi, 1160.
Bloch, Jiles, La eharrue vedique,
411-18; L’lndo-Aryen dn Veda nux
Temps Modenus, 2(13.
Bloomfield. M.. F. Edgerton, and M. B.
Emeneau. Vedie Fnriuaf.''. Vol. III. 213.
Bodde, Derk (translated by). Annual
Customs and Ftstivals in Peking, by
Tun Li-Ch'in, 1132.
Boilding. P. O.. .4 Hanlal Dictionary, Vols,
11, i'll, 222,
Brahmi script, 77 et seq.
Braluii, 981 et seq.
Brandt. J. J., Modern Newspaper Chinese,
1139.
Bray, Denys, The. Brahui Language,
Parts 11, 111, 223.
Brouner, Walter Brooks, and Fung Vuet
Mow, Chinese Made Easy, 1140.
Buddhism, 77 et seq. ; 619 et seq. ;
1013 et seq.
Buddhistic San.skrit, 501 et seq.
Biihler, G. (edited by), Apastambiya-
dhanna.sutram. Third edition, 216.
Burma, alphabets and phonology in, 517
et seq.
Bi rrow, T., The dialectical jxisition of
the Niya Prakrit, 419-43.5.
Burusliaski, 627 et seq.
Bi TLix, R.ay.iioxi) T., On the alphabetic
notation of certain phonetic features of
Malayalam, 437—447.
C
G.aliphate, history of, 143 et seq
Cameron, George C., History of Early
Iran, 1155.
Carell, Grammatical categories in
the New Hebrides, 189-202.
Cauensira, Ease. 11, 239.
Caucasus, history, 613 et seq.
Central Asian documents, 789 et seq.
11
INDEX
Charulra-kanta Abidhan. A comprfhemive
dictionary of the Assamese Innyuage, 235.
Change of a to e in the Indo- Aryan loan
words of Malayalam, The, 55^562.
Changsha, 1055 et seq.
Chabpextieb, Jabl, P6st(a), 101-6;
Sakadhuma, 449-456.
Charrue vedique. La, 411-18.
Chattebji, 8i7xm Kumar, Purana
legends and the Prakrit tradition in
new Indo-Aryan, 457-466 ; A Roman
, Alphabet for Indian 212.
Chiang Yee, The. Chinese Rye. 11.34.
Chinese, logic, 1013 et seq. ; philology,
10.55 et seq.
Chinese Art. An 1 niroductory Haiulbook,
11.34.
Chinese manii.scripts in the fjtein
Collection, 11, 1-26.
Chintamani, T. R. (edited by). The
Urthdisutras with the Vrtti of Rreta-
lanamsin, 217.
Chbistensex, Abthub, Die Iranier, 240 ;
Some new Awromani material pre-
pared from the collections of Age
-Meyer Benedictsen, 467-476.
Codices Avestici et Pahlavir.i Bibliothecae
Universilatis Hafniensh, Yot. V, 11.53.
CoUinder, Bjorn, Indo-Vralisches
Sprachyut, 227.
COOMARASWAMY, AsAXDA K., Two
Vedantic hymn.s from the Siddhdnta-
muktdinU, 91-9.
Copper plate grant of Sriviraraghaya
Cakravartin. 95.5-967.
Cremona. A., Il-Fidwa tal-bdiewn, 1175.
CcxY, A., Les nasales en fin de mot en
Sanskrit (et latin). 477-486.
D
ad-Damiri, 143 et seq.
Darby, A., A Primer of the Marathi
Lnngvage, 221.
Dasgupta, .Jay.anta Kumab, .Some early
dramas in Bengali, 113-1.5.
Dated Chinese manuscripts in the Stein
Collection, 11, 1-26.
Davis, Frank, Chinese., Jade, 1139.
Debrunxer, Albert, Der Typus tvdd-
im Altindischen. 487-.500.
Delaporte, Louis. Byllahnire Hittite
Ctineiforme. 238.
phola-Marura Duha, 97.5-9.
Dialectical position of the Xiya Prakrit,
The, 419-435.
Doke, Clement AL. Bantu Linguistic
Terminology, 1190.
Dravida languages, “ Uralian ” element
in, 751 et seq.
:^ayidian pre6xes, 813 et seq.
Diiraezil, Georges, Rludes com pa
sur les langnes caucasienm ^ ihi
Quest (Morphologie), 239.
Dunne, J. Hey worth (etlited byi. .1'
ar-Rddi wal-Muttaki from th^ A
al-Awrdk, 11.57.
Ihirie, R. V. Dewar-, and -J. B. Davi i
Houston, Chinese and EmjliJi .1/
Military Dictionary, 1141.
Dvandvas avestiques, Sur ((luli
40.5-9.
E
e and o of LuGanda and the n oi Sv.il^
The, I121-.3.
Edgebtox. Fbaxklix. The Prak’
underlying Buddhistic hybrid Saii'-ki
.501-,5I6; (with M. Bloomfield a
.M. B. Enieneau), Vedic Vaniint'. 1
in, 21.3.
Egypt, The .Arabic theatre in. I. I
173-187; 111,991-1012.
E.MEXEAr, M. B.. Phonetic ohservatii'
on the Brahui language, 981-3 ; (w"
M. Bloomfield and F. Edgertci
Yedic Variants, Vol. Ill, 213.
F
Famigia. Dun Karm, TaffhUm -Yi-ou
fuq il-Katekismu fal-Kardinal /
Oasparri, 1178.
Fawcett, Sir Charles, The First t.'infni''
of British Justice in India. 267.
Firth. .1. R., Alphabets and phoiiolo!-''
in India and Burma, 517-546; I'vith
B. B. Rogers), The structure of tic
Chinese monosyllable in a Hiinane-'
dialect (Changsha), 105.5-1074.
Fitzgerald, C. P., China. .4 short cult n rut
history, 1126.
Florida, A grammar of the language ct.
1075-1110.
Forster, Lancelot, The New Culture i”
China, 1126.
French influence on the Arabic theatre,
173 et seq.
Fri.sk, Hjalmar, Zur indoiranisrhen uml
grierhisrhen Nomimlhildung, 214.
Fung Yuet Mow, and AA'alter Brooks
Brouner, Chine.se Made Rn.sy, 1140.
G
0. Bprgstrdsser'tf <
Qmndziige
de<
Lflamischen Rechts^
Bearheitet
ton
, Joseph Sr.harhty 1163-
Qabrata Ward, 1180.
Ga bripl i , Francesco,
II Califfato
di
Hishdm, 1161.
Galea, A. M., S. Gitann Bosro, 1177.
Garratt, G. T., and Edward Thompson,
Rise and Fulfilment of British Rule m
India, 264.
INDEX
III
Gauri Shankar, SrhnarlhhtgavadgUa kd
Poynhkdsdihuvad, 221.
Geiger, Ber>'hard, Mittelpersisch cenok
,,Erbye (Linse ?)“, 547-553.
Geiger, Wilhelm, Singhalesisrhe
Etymologien, 555-7.
Genitif absolu en vieil indien> A propo.s
du, 969-974.
Giles, Lionel, Dated Chinese manu-
scripts in the Stein Collection, II, 1-26.
Gopa Varma, K., The change of a to e
in the Indo-Arvan ioain\or<ls of
Malayalam. 559-562 ; The copper
plate grant of Srivirafaghava Cakra-
vartin. 955-967.
Goodrich, Luther C’arrington. The
Literary Inquisition of Ch'ien-Luny,
1131.
Grammar of the language of Florida,
British Solomon Islands, A, 1075-1 1 10.
Grammatical categories in the New
Hebrides, 189-202.
Gray, Luns H., Observations on Middle
Indian morphology, 563-577.
Grierson, George A., Indian and
Ir.4.nl\n Studies presented to
George Abraham Grierson. 297-
881 : Bibliography of published
writings. 297-318; On the modern
I ndO‘A ryu n Ve) nar u Ut is. 218.
Grousset, Rene, Histoire des Croisodes et
du Royaume Franc de Jerusalem^ //,
243.
H
Hansen, Olaf, Sakische Etymologien,
579-581.
Haslund, Henning, Men and (rods in
Mongolia. 1129.
HayCit al-hayaicdn. A history of the
Caliphate in the, 143-155.
Haydar Mirza Dughlat, Ein turkisches
Werk von, 985—9.
Henning, W., Soghdische Miszellen,
583-8.
Herzfeld, Ernst, Aichnologiva! History
of Iran, 1154; Old-Iranian “peer-
ship", 937-945; iH'r Tod des Kani-
byses ; hvamrsyus amr\'ata, 589-597.
Hcschel, A., Die Prophetic : O Istocie
Prororfwa. 1 166.
Hindi. 363 et seq. ; 373 ct seq.
Hinduism, 107 et seq.
Hjiulustaiu Grammar. 817 et seq.
Hiralal Jain (etlited by), Jayadharald
Tikd of Vlrasen/i, 219 ; Karakam-
dacariu of Kanakdmara Muniy 218;
Pahudadohn of Rdmasimha Muniy 218.
Historia de gramatica eoncani, 715-735.
History' of p in Kanarese, The, 673-683.
History of the Caliphate in the Hnydt
al-hnyairdn of ad-Damirl, 143-155.
Houston. J. B. Davidson-, and R. V.
Dewar-Durie, Chinese and English
Modern Military Dictionary, 1141.
Hsiuiig, S. I. (tran.slated by), The
Romanrr of (he Chamhery 1132.
Himanese <lialect. 1055 et seq.
Hunter, G. K.. Th^ Script of Harappa
and Mohenjodnio and its Connexion
with itther SrriptSy 222.
Hvamr&yus amryata, 589 et ^eq.
Hvatanica, 923-936.
1
li>rahim Maflkour. Ln place <VAl Fdrdhl
dans I'ecole philosophique musulmane.y
2,53.
India, alphabets and phonology in, 517
ct seq.
Indlan and Iranian Studies presented
TO George Abraham Grierson, 297-
881.
Indian dialect.''. Bird-names in the, 599-
601.
Iran, histoiy. 157 ct seq.; sociology,
937 et seq.
Iranian elements in Khouur. 657-671.
Iranian studies, V, 117-142.
Isha q ue, M . . S ukha n va rd n-i-lrdii da r
"a^r-i hdzir. Poets and Poetry of Modern
Pfisia, Vol. /. 254.
Italian influence on the Arabic theatre,
173 ct seq.
IvENs, W. G.. A grammar of the language
of Florida. British Solomon Islands,
1075-1110.
J
Jabagi, M., Textes popidaires Ingus, 239.
Jaina-Mahara^tri, a specimen of archaic,
319-333.
Jainism, 319 et seq.
Japanese, language, 27 et seq. ; 1039 et
seq.
Japanese p<nticles wn.ga. and mo, 27—49.
Ja>ehke. H. A., A Tihetan-Enghsh
Dictionaiy (reprint), 240.
Jenyns. Soame, A Background to Chinese
Painting. 1134.
Joan Josua Ketelaar of Elbing, author of
the first Hindustani grammar. 817-822.
Johnson, F., Knmusi ya Kisivahili yaani
Kitnbu eha Maneno ya Ki-^wahili
kimetungnyf. 1192.
Johnston, E. H., Bird-names in the
Indian dialects. 599-601 ; (with H. W.
Ikiiley), A fragment of the Utfarafunira
in Sanskrit, 77-89.
Johnston, Reginald F., Confucianism and
Modern China, 1129.
*’ Ju-shih Lun ” — a h>gical treatise
ascribed to Vasubandhu, 101.3-1037.
K
Karabyses, Der Tod des, 589-597.
Kanarese, philology-, 673 et seq.
IV
INDEX
Kane, P, \ and S. G. Patnardhan
(translated by). The Vyamharatnu-
yukha of Xilakfintha^ 269.
Karikas of Gaudapada and the
Bhagavata Purana, 107-111.
Kasravi, Sayyid Ahmad, Tarlkh-i pdn-^ad
mla-yi Khuzistan, 1172.
Kastur Mai Bathiya, ydynd-lekha aur
JlunibT, 1150.
Ketelaar, Joan Josua, 817 et seq.
Kharl Boll, 363 et seq. ; 373 et scq.
Kharo^thi inscriptions, 419 et seq. ;
637 et .seq. ; 763 et seq. ; 789 et seq.
Khatak, S. K., and Otto Spies (erlited and
translated by). Three Treatisen on
Mystirism, 1170.
Khotan texts, 923 et seq.
Khowar, Iranian elements in, 657-671.
Konkaui, grammar, 715 et seq.
Koxow, Stex, Note on the ancient
Xorth-Western Prakrit, 6f»3-6I2.
Kramees, J. H,, The military coloniza-
tion of the C'aucaau.s and Armenia
under the Sasaanid.s, 61.3-18.
Manoraohan Ghosh (edited by), Xaii'H- -
Lara's Ahhituiya-darpanam, 217.
March. Benjamin, Some ieehnirnl ti ” •
of Chinese Painting, 1138.
Melanesian languages, 189 et 'c.
1075 et seq.
Middle Indian morphology, 563-.777.
Military colonization of the Caucasus .la .
Amienia under the Sas.sanid'. Tt '
613-18.
Mingana, A. (eriited and translated la
Hook of Treasures. By -JrAt if EP"
250.
Minto, Mary, Countess of. Imtia,
and ilorley, 263.
Mittelpersisch venok ,.Erbse (Linse ,') ■
547-553.
Mleccha, 623 et seq.
Modem literary Hindi, 373-390.
Morgexstierxe. G., Iranian elements n
Khowar, 657-671.
Moutarde, Les noins de la. 703 et seq.
Munda languages, ” I’ralian ’ elemi ii!
in, 7.51 et seq.
L
Lamb, C'orinne, Chinese Fesliie Board,
11 . 33 .
Lamotte. Etienne (edited and translated
by), Sarndhinirmocana Sutra, 1157.
I.,atin, philologj-, 477 et seq.
r^e Fevro, Georges, An Eastern Odyssey,
Ltvi, Sylvaix, Mala vihara, 619-622.
Lierich, B., Xochmals mleccha. 623-6.
Lin \ utang. My Country and My People,
1125 .
Linguistic p.sychologv, 189 et ."eq.
Loc'kh.art, L.ai' rexc'E, Xadir Shah^.s
campaigns in ‘Omrni, 1737-1744, 1.57-
171.
LoRrMER, 11. L. R., Nug.ie Biiruslmskicae,
627-6.36.
Lvoers, HErxRierr. Zur Schrift unri
Sprache der Kharo.sthi- Ihrkuniente,
637-655.
LuGanda, The e and o of, 1121-3.
Lyde, L. \V^., The idoniment of Asia, 266.
M
Mahmoud Mohtar-Katirjoglou, Jm
. mijes.se roranir/ue eclairce par des
rersets choisis, 242.
Maillart. Ella K.. Turkestan Bobs, 260.
Mala vihara, 619-622.
Malayalam, (.opper [jlate grant in, 9.5.5 et
^q. ; Indo-.Aryan loan words in. 559-
56-, On the alphabetic notation ol
certain phonetic features of, 437-447.
Malekula, linguistic psychology in, 189 et
seq. ‘
contribution to
the bibliography of Tilret, 51-76.
X
f . .
Xadir .Shahs campaign.s in ’Oman.
1737-1744, 1.57-171.
Xakasi.mhia, .A. X., The history of p ic
Kanarc.se, 67.3-683.
Narottam Has. Ram .Slh, and Sfirai
Karan (edited by), I)hold-Mdrii m
Duhd, 1147.
X'asaic.s cn fin de mot cn .samskrit (et
latin), 1a>s, 477-486.
Xeue .singhale.sisehe J,autrt'gel, 701-2.
Xew Hebrides, grammatieal eategorie.s in
the, 189-202.
Xiel.sen. K., Enpj>i.sk-Oidhok — Lapp
Dirtionanj, f’id.s. /, If, 234.
XiTTi. Luigia, Grammariens tanlifs et
dialeetes (In pr.ikrit. 681-3.
Xiva in.seription.s, numerals in the.
763 et se(p
Xiva Prakrit, 419 et serp ; 60.3 et ceq. ;
76.3 et stup
Xo<.hmals nderrha, 623—6.
X'oins de la moutarde et (hi sesame, Lcs,
703-8.
Xote.s and (Queries, 1 195.
Xugae Burusha.skieae, 627-6.36.
Numerals in the Xiva in.serijitions. The,
763-779.
O
Gbituary Xotiees —
Chamberlain. Professor Basd Hall, 284.
flharjientier. Professor .Jarl, 282.
Gomtioez, Profe.s.sor Zoltan, 286.
Werner, Professor Alice, 281.
Oertel, Haxxs, The expressions for
'■ The year consists of twelve months ’
an(l the like in Vedic pro.se, 68.5—694.
INDEX
V
Old-Tranian “ peership 937-945.
O’Malley, L. S. S., lndi(i'-'< Social Heritaye,
265; Popular Himluism, 1147.
‘Oman, Nadir Shah's campaigns in,
157-171.
Opium, 101 et seq.
Padm Narayan Acarsa and fShyain
Snndar Das, Bham Rahasya^ Part /,
1151.
Pahlavi, 391 et seq. ; 547 et seq.
Pali acrhati and Sari.sknt m mudern
Indo-Arvan, 795-S12.
Pali bhunaha, 713-14.
Pali, philology, 501 et seq. ; 713 et seq. ;
781 et seq. ; 795 et seq.
Pamir languages, 875 et seq.
Pandit, B. S., Syntax of tlie pa^t ten.se
in Old Rajasthani, 695-8.
Parthischer Titel im Sogdisehen, Ein,
737-749.
Patwardhan, S. G., and P. V. Kane
(translated by), 7'he Vyavaharama-
yukha of yUakantha, 269.
Perham, Margery (edited by). Ten
Africans, 1193.
Perry, E. D., A Satiskrit Primer, yetc
edition y 1153.
Persia, history, 589 et seq. ; 047 et seq.
Phonetic observations on the Brahfu
language, 981-3.
Phonology in India and Burma, 517-546.
Pines, Salomon, Beitrdge zur Islamischcn
Atomenlehre, 1165.
PiSANi, Vittore, Vedico yuh ** ^e
ipsum ”, 699-700.
Ponder. H. W., Cambodian Glory, 1143.
Post(a), 101-6.
Prakasananda, 91 et seq.
Prakrit, grammar, 319 et set). ;
undhology, 457 et seq.; philoU»g\,
603 et .seq. ; 6S1 et se<j
Prakrit underlying Buddhistic hybrid
Sanskrit. The, .501-516.
PRINTZ, 5Vu.helm. Neue singhale.sische
Lautregel, 701-2.
Przyll'ski, J., and C. Kegamky,
nom.s do la nioutarde et du .seNime.
703-8.
Psaila, Mgr., ('‘Dun Kann ”), Vyo
Foscolo. L-Oqbia, 1176.
Purana legends and the Prakrit tradition
in new Indo-Aryan, 457— 166.
Rajasthani, 695 et seq.
RajputanI, poetry, 975 et seq.
Ram Nares Tripathi (edited by), Tula
Dds's Rdmcaritmdnns {Rdmdyan), 1148.
Ram Sih, Suraj Karan, and Xarottam
Das (edited bv), Dhold-Mdrurd Duha.
iU7.
Ramaprasad Chandia, Medieval Indian
Sculpture in th^ British Musputn, 1146.
Rapson, E. J., Sanskrit sd and sdh,
709-711.
Ray, Amarnath. Bhagavata Purana and
the Kiinkas of Gaudap.ida, 107-111.
Regamey', C., and J, Przyllskj. Les
noma de la moutarde et du sesame,
703-8.
Remarks on the Pahlavi ligatures
Hevifus (»f '20‘S ; 1J25.
Heiuf lUi Elu/Uf! Islutniqui Tom'- VIII,
241 ; Tontr IX, ll(>4.
Rogers, i5. J!.. ami J. H- Firth. The
structure of tiie Chinese monosyllahle
ill a Hunanese dialcit (Changsha),
J0.5.'.-i074.
Rooke. (i. 4i. (trausiated by), The.
Meijhmlutn of Khliilh’'!, 1152.
Ro.ss. Sir E. R.. and Rachel O. Wingate
(oollecteri and edited by). Dialogues in
the Eii'^teni Turli Diah-rt on Subjects of
Interest to Trarellere, 2')7.
Ruaka, Julius (eilitcd and translated by),
Das Such der Alaune und Hake, 252.
Saka, 117 et .«eii. : 579 et .seq.
Sakadhuma. 449—456.
Sakisehe Etymologien. .579-.581.
Saksex. 4. BABVR.yM, Pali bhUunhn, 713-
714.
Saldanha. MABi.y:NO, Hi.st6ria de
gramatica oonrani, 71.3-7.3.5.
Sanskrit d-kseti and Pali archnti in
modern Inilo-Arvan, 79.3-812.
Sanskrit, grammar, 7U9 et seq. ; 9(19
et seq. ; philology, 411 et seq. ; 449
et seq. ; 477 et seq. ; .301 et seq. ;
623 et .seq. ; 795 et seq. ; philosophy,
77 et seq. : 91 et seq. ; 619 et seq. ;
religion, In7 et .seq.
Sanskrit su and sdh, 709-711.
Sassaiiids, military eolonization by,
61.3 et seq.
Satisehamlra Seal and Sitanath Pradhaii
(edited by). Hgreda-Hamhitd, Part I,
21.3.
Saunders, Kenneth, The Idetds of East
and IFfst, 1128; A Pageant of Asia,
112S.
Saussey, E., Prosateuis turcs contem-
parains, 2*39.
Savdon, P. P., and Jos. Aquilina, Ward
fa' Qati Malti, 1176.
Sbath. Paul (edited by). Trade's Rtligitux,
Philosnphigues et Moraux, Extraits
des (Einres d'Isiiac de Xinive, 1167.
Sehaeht, Joseph (edited by). G.
Bergstrdsser s Grntukuge des Islam-
ischen Rechts, 1163.
VI
INDEX
kScHAEDEB. Haxs Heinrich, Kill
parthischer Titel ini .Sogdi.schen, 737-
749.
.Schmidt, t'arl (edited by), Manichai^he
Hundsckriflen der StaatUehen, ilmeen
Berlin, Band I, 238.
.SCHRAIIER, F. Otti), On the • I'raliaii ”
element in the Dravida and the
Murida language..^, 7.31-762.
Schwab, Raymond, Amjaelil-J/uperroii,
sa Vie, 26.3.
.Sesame. Res noms dii, 763 et seq.
Sh.vam Snndar Dii.s and Padni Xar.Iyan
Acarya, Shd^d-Bahanya, Part /, 11.51.
Siddhdnifnnnktdrnli, Wdantie hvmns
from the, 91-9.
Silcock. Arnohl, I nltadnctifni to k'kinene
Art, 1134.
Singhale.si.s< he Et\ niologien, 5.5.>-7.
Sinhalese, .3.3.5 et seq. ; 701 et seq.
Siren, Osvald (translated by). The Chinese
on the Art of Paintim/, 1 1 37.
iSitanath Fradhan and Satisehandra Seal
(edited by), Jirjreda-Samhdn , Part ],
21.3.
Skold, Haniies, Materialien zu den
Iranhrhen Pamirrprarhen, 1)36.
Smith, Margaret, ,4a Early Myntie oj
Baghdad. 1168.
Soghdian, .38.3 et .seq. ; 7.37 et seq,
.Soghdische Miszellen. .38.>-8.
.SoMOGvr, .Joseph he, A History of the
Caliphate in the Haydt aUhayaartn of
ad-Damlri. 143-1.3.3.
.Spio.s, Otto, and .S. K. Khatak (editeil
and tran.slated by). Three Treatiie.s on
Myetiehm, 1170,
iSri Uniesa Misra (edited bv), Kr.^na-
jatima of Manhodh, oitt,
6rl\ Ifafaghava Cakravartin, copper plate
grant of, 95.3-967.
.Stein Collection, Dated Chine.se niaiui-
scripts in the, II, 1-26.
Stein, O., The nunieral.s in the Niya
Inscriptions, 76.3-779.
.Storey, C. A., Persian Literature, Section
fl. Ease. 1, 2.3.3.
.Structure of a Rantii language with
special refen.nee to Sw.ihili, The
1111 - 1120 .
.Structure of the Chinese monosyllable
m a Hunanese dialect (Chang.sha). The
10.3.3-1074.
Suraj Karan, Xarottain Da.s, and Riim
.sih (edited byl. Dholf,- Marfira m/id.
Sushil Kumar De (edited by), 3
PadynralT of Papa (lonramin, 217.
.Sutcliffe, Edmund, A dranirnar of
Maltese Language, 1180
Swahili, nil pt seo . li.ri »
Syntax of the past teii'C m Olt
Rajasthani, 695-8.
T
Tahdgaya and Tathagata, 781-8.
TAglZ-ADEH, S. H.. .4 new contribution t.
the materials concerning the life o'
Zoroaster, 947-954.
Taraporew'ala, I, .J. .S., SeJ<rtions fo’/r
Classical Gujarati Liiteratiire, Vol. II.
221 .
Tathagata and Tahagaya, 781-8.
Thomas, FI. J., Tathagata and TahJtr.i' .i,
781-8.
Thom.xs, F. \V'.. Some words fouiul in
Central .Asian documents, 789-794.
Thompson, Edward, and G. T. Garnitt.
Rice and Fulfilment of British Rale in
India, 264.
Tiliet, hihliography, 51-76.
Title.s and .Summaries of The.«es, 287.
Toeharistan, 883 et seq.
Tod des Kamhvses : hvararav us anirv.d.i.
Der, .389-597.
Tiala, Kenji, and Lucy Driscoll, Chtnt'i
( 'ailiijra phy, 1142.
Torwali. re.semblanoe.s of Xi\a Prakrit to,
419 et seq.
Ttaugara, 883-921.
Tilda- im .Altindi.schcn, 487-.300.
Tun Li-Ch'en, Aniiiia! Customs and
Fe.stirals in Peking, 1132.
Tnrfan, a Fahlavi-frahang from, 391 et
seq.
Turki-sehes Werk von Ha.vdar-Alir/ia
Dnghlat, Ein, 98.3-9.
Ti knek, R. L., .Samskrit d-kseti and Pah
archati in modern lndo-.4ryan, 795-
812.
Tittle, Edui.x H., Some Dravhli.aii
prelixes, 813-15.
U
" Uralian element in the Dravida and
the .Mund.i language.s. The, 751-762.
I'rdn grammar, 817 et seq.; philology,
363 et seq.
Vtlaratnn.tra, a fragment of the, 77-89.
latkiallo, A. V., Dell ta’ l-imglioddi, 1177.
\ .vssiLiEv, Boris, “ Ju-shih Enn ’ —
a logical treatise n.scrihcd to \'a.sii-
bandhn, 1013-1037.
Xasubandhu, a logical treatise ascribed
to, 1013-1037.
\ asiidevahiijdi, a specimen of archaic
Jaina-.VIaijar5.stri, The, 319-333.
V edantie hymns from the Siddhdntn-
muktdvalT, 91-9.
Wdie, philology, 411 et seq. ; 449 et
seq. ; prose, 685 et seq.
Vedico yuh “se ipsum", 699-700.
INDEX
VU
Venok, 547 et seq.
Vogel, J. Ph., Joan Josua Ketelaar of
Elbing, author of the first Hindustani
grammar, 817-822.
Vostrikov, Andrew, Some corrections
and critical remarks on Dr. Johan van
Manen’s contribution to the biblio-
graphj' of Tibet, 51-76.
W
Wackernagel, j., Altindische und
mittelindische Miszellen, 823-834.
Waley, Arthur, The Way a>ul itx Poii’er.
A study af the Tao Te Ching, 1130.
Ward, Ida C., An, Introduction to the Ibo
Language, 1188.
Wauchope, R. S., Buddhist Cave Temples
of India, 1145.
Willmak-Grabowska, H. de, a propo.s
du genitif absolu en vied indien,
960-974.
Wingate, Rachel 0., and Sir E. D. Ross
(collected and edited by). Dialogues in
the Eastern Turki Dialect on Aiibjecls
of Interest to Travellers, 257.
Woolner, A. C. (translated by). The
Jasmine Garland {Kundamdla), 1152.
Wortkundliche Beitrage zur arischen
Kidturgeschichte und Wclt-Anschau-
ung, II, 835-873.
W UST, W.ALTHER, Wortkundliche Beit-
rage zur arischen Kulturgeschichte und
Welt-Anschauung, II, 835-873.
Y
Yazdi, 33.5-361.
Yazghulaml texts, 875 et seq.
Yosiiitake, S., The Jap.ancse particle.s
iva, ga, and mo, 27—49 ; A new classi-
lication of tlie constituents of sjwken
Japanc.sc, 1039-1053 ; The Plmnetic
System of Ancient Japanese, 271.
Yuh, 699 et seq.
Yiisuf ’Ali, .4.. Tim Holy Quran. Ejiglish
text and commentary {with Arabic
tert). Part I, 242.
Z
Zarubin, I. I., Two Yazghulaml texts,
87.5-881.
Ziy.I Muhammad. Yiidgar i Tarii, 1149.
Zoroaster, life of, 947 et seq.
lO'-
Stephen Austin and Sans, Ltd., Pri/ders, Hertford.
D.G.A. 80.
CENTRAL ARCHAEOLOGICAL LIBRARY
NEW DELHI
Call No.—
Author — 4 5*// 7
Title—
Borrower's Name Date of Issue Date of Return
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11